Skip to main content

Full text of "Ancient Egyptian Chronology"

See other formats

Ancient Egyptian 

Edited by 

Erik Hornung, 

Rolf Krauss, and 

David A. Warburton 


Ancient Egyptian Chronology 



Ancient Near East 

W. H. van Soldt 


G. Beckman • C. Leitz • B. A. Levine 

P. Michalowski • P. Miglus 

Middle East 
R. S. O'Fahey • C. H. M. Versteegh 


Ancient Egyptian Chronology 

Edited by 

Erik Hornung, Rolf Krauss, and David A. Warburton 





This book is printed on acid-free paper. 

Library of Congress Cataloging-in-Publication Data 

Ancient Egyptian chronology / edited by Erik Hornung, Rolf Krauss, and 
David A. Warburton; with the assistance of Marianne Eaton-Krauss. 
p. cm. — (Handbook of Oriental studies. Section 1, The Near and 
Middle East ; v. 83) 
Includes bibliographical references and index. 
ISBN- 13: 978-90-04-11385-5 
ISBN- 10: 90-04-11385-1 
1. Egypt — History — To 332 B.C. — Chronology. 2. Chronology, Egyptian. 
3. Egypt — Antiquities. I. Hornung, Erik. II. Krauss, Rolf. III. Warburton, David. 
IV. Eaton-Krauss, Marianne. 

DT83.A6564 2006 
932.002'02— dc22 


ISSN 0169-9423 

ISBN- 10 90 04 11385 1 

ISBN- 13 978 90 04 11385 5 

© Copyright 2006 by Koninklijke Brill NV, Leiden, The Netherlands. 

Koninklijke Brill NV incorporates the imprints Brill, Hotei Publishing, 

IDC Publishers, Martinus Nijhoff Publishers, and VSP. 

All rights reserved. No part of this publication may be reproduced, translated, 

stored in a retrieval system, or transmitted in any form or by any means, 

electronic, mechanical, photocopying, recording or otherwise, without prior 

written permission from the publisher. 

Authorization to photocopy items for internal or personal use is granted 

by Brill provided that the appropriate fees are paid direcdy to 

The Copyright Clearance Center, 222 Rosewood Drive, Suite 910, 

Danvers, MA 01923, USA. 

Fees are subject to change. 



With this volume of the Handbook of Oriental Studies (Handbuch der 
Orientalistik) a new editorial board for the ancient Near East takes 
over from the board that has edited the HdO volumes for so many 
years. The new editors have been chosen because of their expertise in 
the fields that pursue the study of the civilizations of the ancient Near 
East. These editors are: G. Beckman (Hittitology), C. Leitz (Egyptology), 
B. Levine (Hebrew Bible/Ancient Israel), P. Michalowski (Sumerology), 
P. Miglus (Near Eastern Archaeology), and W. H. van Soldt (Assyriology, 

The board has planned a number of new volumes in the HdO 
series and expresses the hope that with these volumes the series will 
resume its original function of a handbook for Ancient Near Eastern 
Studies. At the same time, the new board is committed to the publi- 
cation of volumes that had already been planned and accepted and 
which are scheduled to appear in the near future. However, it should 
be pointed out that the present board members do not necessarily agree 
with the entire content of a volume that they have 'inherited' from 
their predecessors. 

Finally, we would like to thank the members of the former editorial 
board for all their work for the HdO series. 

The Editors 



Introduction ... 

Erik Hornung 



1. Royal Annals 19 

The Editors 

2. The Royal Canon of Turin 26 

Kim Ryholt 

3. King Lists and Manetho's Aigyptiaka 33 

The Editors 

4. Genealogy and Chronology 37 

Morris L. Bierbrier 

5. Methods of Dating and the Egyptian Calendar 45 

The Editors 


1. Predynastic — Early Dynastic Chronology 55 

Stan Hendrickx 

2. Dynasties 0-2 94 

Jochem Kahl 

3. Dynasty 3 116 

Stephan J. Seidlmayer 


4. Dynasties 4 to 5 124 

Miroslav Verner 

5. Dynasties 6 and 8 144 

Michel Baud 

6. First Intermediate Period 159 

Stephan J. Seidlmayer 

7. Middle Kingdom and the Second Intermediate Period 168 

Thomas Schneider 

8. New Kingdom 197 

Erik Hornung 

9. Dynasty 21 218 

Karl J arisen- Winkeln 

10. Third Intermediate Period 234 

Karl J ansen- Winkeln 

11. Saite and Persian Egypt 265 

Leo Depuydt 

12. Nubian Kingdoms, Dyn. 25 through the Kingdom of 

Meroe 284 

Karola ^ibelius-Chen 

13. Chronological Links between the Cuneiform World of the 
Ancient Near East and Ancient Egypt 304 

Jo'rg Klinger 


1. Radiocarbon Dating and Egyptian Chronology 327 

Start Manning 

2. Luminescence Dating of Egyptian Artefacts 356 

Christian Goedicke 

3. Dendrochronology 361 

Otto Cichocki 

4. Dates relating to Seasonal Phenomena and Miscellaneous 
Astronomical Dates 369 

Rolf Krauss 

5. Astronomy on the Horizon — A Tool for ancient Egyptian 
Chronology? 380 

Juan Antonio Belmonte 


6. Lunar Days, Lunar Months, and the Question of the 

Civil based Lunar Calendar 386 

Rolf Krauss 

7. Long-term Variation in the Motions of the Earth and the 
Moon 392 

Kurt Locher 

8. Lunar Dates 395 

Rolf Krauss 

9. The Heliacal Rising of Sirius 432 

Teije de Jong 

10. Egyptian Sirius/Sothic Dates and the Question of the 

Sirius based Lunar Calender 439 

Rolf Krauss 

11. Foundations of Day-exact Chronology: 690 BC-332 BC .... 458 

Leo Depuydt 


1. Conclusions and a Postscript to Part II, Chapter I 473 

Rolf Krauss & David A. Warburton 

2. Chronological Table for the Dynastic Period 490 

The Editors 

3. Tables for Kushite Rulers 496 

Karola 2jbelius-Chen 

Abbreviations & Bibliography 501 

Index of Personal Names 509 

Index of Place Names 514 

General Index 516 


It was Wolfgang Schenkel who suggested the project of a Handbook of 
Egyptian Chronology to Brill, Academic Publishers, nearly a decade ago. 
He will have asked himself in the intervening years whether the work 
would ever appear. 

Co-editor Erik Hornung developed the concept of the Handbook 
and together we selected the authors for each section. The subsequent 
addition of David Alan Warburton to the team provided an impetus 
to overcome the inertia that had set in. He and Marianne Eaton-Krauss 
took over the thankless task of translating and/or editing contributions 
by authors whose mother tongue was not English, except for Chapters 
II. 1 and 11, III. 7, 9, and 11. 

I am indebted to the Stiftung Preussischer Kulturbesitz who trans- 
ferred me from the Agyptisches Museum to the Museum fur Vor- und 
Fruhgeschichte in 2001 to do research on the Calendrics and Chronology 
of ancient Egypt, the Near East and Europe. Although the move 
deprived me of direct access to an Egyptological library, the Handbook 
would not have seen the light of the day without this change of for- 
tune. Thanks to Wilfried Menghin, Director of the MVF, I could con- 
centrate on the Handbook during 2004 and 2005. 

Last but not least, I thank Juan Jose Granados who encouraged me 
to complete the unloved task. 

R.K. February 2006 


Erik Hornung 

It is widely accepted that chronology is the "spine" of history. The 
ensuing image is slightly distorted since one can only extend or shorten 
a spine using violence, stretching it or forcing it like the giant Procrustes 
into a bed which is far too short. In recent years there have been 
repeated attempts to cut an entire century off history, and not merely 
in the "Dark Ages" of antiquity, but even now for the supposedly 
"invented" Middle Ages. It would thus be more suitable to view time 
as the clothing of history: a garment can be non-violently adjusted to 
a changing body. The image of time as a piece of clothing, the habit 
of mankind and the dress of the deity, can be found in the poem "Song 
of Time" by the Pakistani poet Muhammad Iqbal (1877— 1938), ' who 
was doubtless inspired by Goethe's expression, "the living dress of the 
divinity" created by the spirit of the Earth "on the whirring loom of 
time" (Faustus, verses 508f). 

But, first a word on the temporal horizon of the Egyptians. 2 Even 
'Ankhtifi of Mo c alia looked hopefully to a future of "millions of years", 
during which no one would approach his deeds. 3 Indeed, before the 
end of the FIP, these "millions of years" rapidly became a common- 
place (being known from a wish for King Merykare c in Asyut), 4 ulti- 
mately becoming a synonym of the twin temporal concepts neheh and 
djet. This is most clearly expressed in the representation of the Dyn. 
2 1 illustrated papyrus of Khonsumes in Vienna, where all three stand 
together, raised on yokes and thus divine, 5 subordinated to the solar 

1 In A. Schimmel, ed., Botschaft des Ostens (Tubingen & Basel, 1977), 149. 

2 E. Hornung, "Zeitliches Jenseits im alten Agypten", Eranos-Jahrbuch 47 (1978), 
269-307, esp. 297-298. 

3 Schenkel, Memphis, 47. 

4 F. LI. Griffith, The Inscriptions of SOU and Der Rifih (London, 1889), pi. 13, IV 22. 
In Dyn. 6, Pepy II was wished "millions of ^-festivals" (Urk. I, 115, 1). 

' H. von Demel, "Der Totenpapyrus des Chonsu-Mes", Jahrbuch der Kunsthistorischen 
Sammlungen in Wien, N.S. 13 (1944), 1—16; A. Piankoff& N. Rambova, Mythological Papyri 
(New York: BS XL, 1957), no. 16; djet does not appear as a divinity, but rather form- 
less, expressed in hieroglyphic writing. 


orbit which constantly regenerates time while the sun itself is in the 
tow of the hours: solar circuit and time are thus indivisibly woven 
together. The memorial temples of the NK Pharaohs are thus "Houses 
of Millions of Years" as they guarantee the mortuary cult for the rest 
of the horizon of time 1 ' just as the tomb is the place of "the fullness of 
time (neheh)." 7 The limited individual stream of life flows into this sea 
of the "millions of years", as Pharaoh ends his in "millions of sed- 
festivals" in a specifically royal unit of time. In typical moderation, 
Akhenaten wishes to be buried in the new residence Akhetaten "after 
the millions of sed festivals which my father Aten has envisioned for 
me", while Nefertiti and the princess Meritaten are to be buried "after 
millions of years". 8 

In the Ramesside era, the horizon of the future grows even further, 
to "millions of millions of years", and thus exceeding our own astro- 
nomical temporal horizons. 9 The Inscription Dedicatoire in Abydos expresses 
the Ramesside spirit: in wishing Ramesses II "millions more of years 
than the lifetime of Re in heaven", 10 it thus exceeds the temporal dimen- 
sions of the cosmos itself. The Egyptians probably had the feeling that 
there was immeasurable time in the Beyond which was yet not beyond 
the bounds of time. The dead do not belong to eternity, but to the 
fullness of time benefiting from a new lifetime with every nightly pas- 
sage of the sun-god. In the Beyond, however, all past time is accu- 
mulated in the "Place of Annihilation" (hetemit): it is there that the hours 
fall when they have passed, when they are "swallowed" by the time 
which "gave birth" to them. 

The greatest possible extreme appears in the post scriptum to Book 
of the Dead, chapter 62, promising the deceased (in the role of the sun- 
god Re!): "The fullness of time (neheh) without limits is given to me, 
for I am the heir of Neheh to whom djet is given". One immediately 

b Cf. M. Ullmann, Kijnig fur die Ewigkeit — Die Hduser der Millionen von Jahren: eine 
Untersuchung zu Konigskult und Tempeltypobgie in Agypten (Wiesbaden: AAT 51, 2002). 

7 References in P. Vernus, "La stele C 3 du Louvre", RdE 25 (1973), 217-234, 
esp. 223-224 (i), also the occasional designation of the Necropolis. The tomb can also 
be the "Horizon" of neheh, cf. F.-J. Schmitz, Amenophis I. (Hildesheim: HAB 6, 1978), 

8 Murnane & Van Siclen, Stelae, 25, 41. 

9 The statue of Bakenkhons in Munich, KRI III, 298; Hymn in P. Berlin 3049: 
J. Assmann, Agyptische Hymnen und Gebete (Zurich & Munich, 1975), 127 B. A first time 
on the Cairo stele CG 34025 of Amenhotep III: Urk. IV, 1653, 16. 

10 Inscription Dedicatoire, line 27 = 31; KRI II, 325, 7-8; cf. U. Luft, Beitrage zur 
Historisierung der Gotterwelt (Budapest: SA 4, 1978), 173. 


recalls the dialogue of the Beyond in chapter 175 where Osiris asks 
Atum about the lifetime in the Beyond and is giving the comforting 
assurance that it is "millions of millions of years". 

This concerns the future. As far as the horizon of the past is con- 
cerned, the Royal Canon of Turin 11 gives totals for the dynasties of 
the gods and demi-gods ("The Followers of Horus") before Menes and 
the beginning of the Dynastic Period which take us to nearly 37,000 
years. The Early Dynastic kings are assigned long reigns, totalling almost 
another 1 000 years up to Izezi, so that the Egyptians of the NK could 
look back towards a temporal horizon of nearly 40,000 years since the 
start of the world. That is a great deal when compared with the figures 
in the Christian and Jewish traditions. This matches statements made 
by the ancient authors. Herodotus (2, 142) reports that the Egyptian 
priests accounted for a human line of ancestors of 341 generations, 
which he reckoned amounted to 11,340 years before which lay the age 
of the gods. Greek historians, such as Hekataios of Abdera already 
reached the gods in the 16th generation. Genealogies as long as those 
given by Herodotus are generally encountered in the LP, the best known 
being the family tree of Memphite priests of Dyn. 22 which includes 
more than 60 generations. 12 In another note, Herodotus (2, 145) dates 
Hercules 900 years before his own day, and Dionysos "some 1000 

Our principal witness for the ancient Egyptian chronology is Manetho, 
and he must likewise have had similarly high numbers which were then 
forced into a truly Procrustean bed by the Christian authors who thus 
made it compatible with the Biblical chronology whose priority could 
not be doubted. At the same time, however, some individual items were 
systematically extended in order to achieve accord (particularly for the 
period between Menes and Adam), or in order to make the Egyptian 
past appear older. 13 Even in pre-Christian times, there were efforts to 
claim a chronological precedence for Greece at the expense of Egypt. 
Eudoxus of Cnidus (d. 356 BC in Egypt) opposed Herodotus' figures 
using the trick of transforming years into months. 14 From Manetho, 

11 Gardiner, Canon; KRI II, 827-844. 

12 Berlin 23673: Borchardt, Mittel, 96—100; a good illustration of the document will 
be found in the catalogue A. Grimm, S. Schoske, D. Wildung, Pharao: Kunst und Hemchqft 
im alien Agypten (Munich, 1997), no. 89. 1. 

13 Gf primarily Helck, Manetho, 76, 83; cf. also Krauss, Amarnazeit, 239 and Beckerath, 
Chronologie, 38. 

11 Cf. S. M. Burstein, "Images of Egypt in Greek historiography," in: Antonio 
Loprieno, ed., Ancient Egyptian Literature (Leiden, 1996), 591-604, esp. 596f. 


Georgios the Synkellos (8th cent. AD) cites a figure of 1 1 ,985 years for 
the reigns of the gods (and thus close to that of Herodotus), with 
Hephaistos = Ptah assigned 9000 years. This Synkellos simply inter- 
prets as months, which allows "only" 727 3 A years for Ptah. From some- 
time after the reign of Ramesses II, this Memphite god had usurped 
the place of the Sun-god Re at the head of the divine dynasty although 
Diodorus (I: 13, 2) still maintains the older tradition (with Helios as 
the first king of Egypt). 15 Eusebius claims that 13,900 years separated 
Hephaestus and "Bites", following which is another total of 11,025 
years, and thus 24,925, which he then interprets as months like Synkellos, 
reducing them to 2,206 years, which fitted perfectly with the Biblical 
2242 years between Adam and the Flood, while radically cutting the 
Egyptian temporal horizon. 

In the traditional chronicle of the 30 dynasties, with 113 genera- 
tions, the total number of years is named as 36,525, whereby the largest 
share of 30,000 falls to Hephaestus; as the last native Pharaoh Nectanebos, 
some 15 years before Alexander the Great, is assigned the Anno Mundi 
5147. In his work on Life and Opinions of Philosophers, written around 
200 AD, Diogenes Laertios goes further: he calculates that 48,863 years 
separated the invention of philosophy by "Hephaestus, the son of the 
Nile" and Alexander the Great — and he adds the number of solar and 
lunar eclipses in this period. From Zoroaster to Xerxes, the Lydian 
Xanthus reckoned 6000 years, and Plutarch gives a similar estimate in 
De hide 46 placing 5000 years between Zoroaster and the Trojan War. 
This "older" Zoroaster, who belongs in the seventh mill. BC, plays an 
important part in modern esoteric literature, frequently as the teacher 
of Hermes Trismegistos. 

The figure of the king "Menes" the Egyptians created a fictional 
beginning for the historical era. With echoes of the gods Min and 
Amun, it also served as a shortened version of the name of Thutmose 
III, and thus served in an ideal fashion as a link between the world 
of the gods and the Dynastic kings. It is not, however, a mere ques- 
tion of the identity of the names as the Egyptians were frequently 
satisfied with distant echoes. Every attempt to equate Menes with a 
concrete name of the Archaic Period, particularly Narmer and c Aha, 
has been fruitless. 16 With the fictive image of Menes, known since the 

15 Immediately following this, however, he allows a variant that, "according to some 
priests" it was Hephaestos. 

16 On this issue, cf. Hornung & Staehelin, Skarabaen, 44—45, and H. Brunner, "Menes", 


time of Hatshepsut and placed at the head of the list of kings for the 
first time by Sety I in the table of Kings at Abydos, the beginning of 
history is given a form, as in Genesis the beginning of humanity is 
attached to Adam as the first human. The Christian chronographers 
perceived the analogy and attempted to bring the two into temporal 
accord. As the founder of the Egyptian state, the deeds of Menes — 
emptying the swamps and founding Memphis — make him a bearer of 

The universal historian Orosius reckoned some 5200 years separated 
Adam from the birth of Christ, and the figures used by the other early 
Christian authors generally fit into the same general framework. According 
to Sextus Julius Africanus (ca. 2 1 7) the Creation dates to 5500 years 
before the birth of Christ. Jerome (Hieronymus, 347—419) gives the 
date of 5198 BC, Victorius of Aquitaine is exact with 25/111/5201 BC. 
The medieval chroniclers, such as the Russian Nestor Chronicle from 
the early 12th century or the continuation of the Chronicle of the 
Frankish kings by Fredegar reckon that the world began around 
5200/5500 BC, while the Jewish Calendar places it slightly later, 
7/X/3761 BC, although the Kabbala of the 13th and 14th centuries 
estimated the duration of the worldly ages at 7000 years each, the suc- 
cession ending absolutely in the 50th millennium. 1 " 

This temporal horizon, with a maximum of six or seven thousand 
years, was maintained through the Renaissance, and even to some 
extent afterwards. The 18th century Freemasons began their calcula- 
tions with the round figure of 4000 BC (which is also found in Zedler's 
Universal-Lexikon), 19 athough in 1704 Pere Pezron made Year 1 of the 
world 5872 BC, and thus substantially earlier; 20 whereby he also assigned 
priority to the Assyrians and Chinese, before the Egyptians. It was at 
about this time that Isaac Newton attempted to "improve" the Egyptian 
chronology, in order to recover the priority for the Old Testament, 
and because the Egyptians "in their vanity had made their monarchy 

LA IV (1982), 46-48; there is also material from Nubian Sai, discussed by J. Vercoutter, 
"A propos des Mni = Menes", in: S. Israeli t-Groll, ed., Studies Lichtheim, II 1025-1032, 
and the link with Memphis by J. P. Allen, "Menes the Memphite", GM 126 (1992), 

17 H. Brunner, "Menes als Schopfer", ^DGM 103 (1953), 22-26. 

18 G. Scholem, Die judische Mystik in ihren Hauptstrbmungen (Frankfurt, 1967), 195. 

19 Johann Heinrich Zedler, Grosses volbtdndiges Universal-Lexikon (Halle & Leipzig, 1 749) 
LXI, col. 818. 

20 Paul Pezron, Defense de I'antiquite des terns (Paris, 1704). 


a few thousand years older than the world". 21 He could not imagine 
that only a century later William Herschel would work with the hypoth- 
esis that the world was two million years old, and that even Kant had 
assumed a substantially greater temporal horizon. 

It is a fascinating and still undescribed phenomenon how the great 
voyages of discovery moved in parallel so that the Spanish, British, 
Portuguese and other seafarers moved forth into hitherto unknown parts 
of the world, bringing new continents into view, and thus the new dis- 
coveries in space went hand in hand with the extension of time, ulti- 
mately exploding the far too narrow temporal confines of the Christian 
chronographers. And, again Egypt played a decisive role since the 
ancient pre-Christian records were again put to use from the Renaissance 
onwards. The decisive quantum jumps which continue up to the pre- 
sent day began in the late 1 8th century as the study of geology began. 
In 1778, Buffon estimated the age of the earth at 100,000 years, a 
figure which was used by Goethe in Faustus when Mephistopheles 
remarks on Faustus's vision of ruling the sea: 

This is naught new for me to explore 

Knowing this a hundred thousand years of yore (verses 1021 Of.) 

Similarly Cuvier, who in an elegant phrase of Heinrich Heine's, "proved 
in the most ungallant fashion that our mother earth is many thousands 
of years older than she had hitherto admitted", 22 and Jean Paul is sur- 
prised that "the earthy sphere . . . grows older by the day, backwards 
(and not just forwards)". 23 It must have been a deeply wrenching expe- 
rience for the times that the firmly established temporal horizon of the 
Creation in the OT would be weakened with such terrifying speed. 

With the temporal requirements of geology demanding ever greater 
spans of time, the 19th century saw the final abandonment of the 
chronology of the OT. 24 In 1870, Lepsius still assumed that the most 
ancient remains of mankind and their worked stones were 30 to 50,000 
years old. 25 However, millions of years were soon accepted; even Herschel, 

21 Isaac Newton, The chronology of ancient kingdoms amended (London, 1728), 191. 

22 Heinrich Heine, Franzdsische ^justande, Artikel VIII, 27/V/1832. 

21 Jean Paul, Der Komet oder Mkolaus Marggraf: eine komische Geschichte (1822), chapter 
VI, note. 

24 The trail-blazing three volume work of Charles Lyell, Principles of Geology (London, 
1830-33), with numerous improved editions. 

25 R. Lepsius, "Uber die Annahme eines sogenannten prahistorischen Steinalters in 
Agypten", <TS 8 (1870), 89-97, 113-121, esp. 90. 


whom Jean Paul cites, reckoned with 2 million light-years for the most 
distant stars, and thus a corresponding age for the universe; Kant refers 
to "millions of years and centuries" in his General Natural History and 
Theory of the Heavens. In the 20th century, astronomy would change this 
to billions of years — whereby we come closer to the Ancient Egyptians. 
In 1929, Hubble's discovery of the expansion of the universe gave 
another push. We have now all experienced an age during which the 
cosmos gets to be billions of light years larger, and consequently older, 
every couple of years. In fact, the process does not appear to be 
finished — a fine example of how rapidly one can adjust to the extreme 
acceleration of exploding time in history. 

After this time travel to the further horizons, we can return to 
Egyptian chronology, as we understand it today. From the beginning, 
Egyptology had an intense preoccupation with chronology, and gener- 
ally took the highest available figures of antiquity, which were once 
again in vogue since the Renaissance — rather than the Biblical figures. 
Champollion-Figeac placed the debut of the Dynastic Period at 5867 
BC, and that of Dyn. 18 at 1822; John Gardner Wilkinson who always 
preferred to orient himself on the OT used the values of 2320 and 
1575; Lepsius lay between them at 3892 and 1591. 26 One can see that 
they rapidly came close to the presently accepted dates for the NK, 
but that the earlier period was still the subject of great differences in 
the 19th and early 20th centuries AD. Petrie consistently maintained 
that the historical period began in the 6th millennium and even in 
1935 Borchardt put Menes at 4056 BC whereas Eduard Meyer, fol- 
lowed by Breasted, came close to our own estimates with 3315. Fun- 
damental for the earlier chronology was the discovery of the Illahun 
Papyri (ended in 1899) with their Sothic and lunar dates which offered 
the first fixpoint before the NK. We can marginally note that even in 
1870, Lepsius (immediately joined by Georg Ebers) energetically dis- 
missed an Egyptian prehistoric stone age, and attributed all of the stone 
tools to the historic period. 27 It was only after 1890 that a window into 
the deeper prehistoric past of Egypt was opened with a series of blows 
beginning with the discovery of the Naqada culture and fundamental 

26 A. Wiedemann provided a useful synopsis of the early propositions in Agyptische 
Geschichte (Gotha, 1884), 732f, and Borchardt, Annalen, 48—51, enlisting their "Main 

27 Lepsius (n. 25); cf Georg Ebers, "Uber die Feuersteinmesser in Aegypten", ^AS 
9 (1871), 17-22. 


reflections of Georg Steindorff. After some initial uncertainty with a 
broad spectrum of variations, the framework was gradually refined in 
the course of the 20th century, and C-14 dates for the OK now lead 
to considerable irritation when they lie a mere century above the other- 
wise acceptable values. 

Egyptian chronology is still the touchstone by which all of the other 
chronologies in the ancient world are measured and the issue of its 
reliability is thus central. A survey and examination of the chronolog- 
ical prospects for the third and second millennia BC in the different 
parts of the ancient world was the object of a series of meetings, begin- 
ning with Gothenburg in August 1987 at the initiative of Paul Astrom, 2 " 
and continued in August 1990 by Manfred Bietak at Schloss Haindorf, 29 
and in November 1996 at the same venue, and then in May 1998 in 
Vienna. These last two already took place under the auspices of Bietak's 
major Sonderforschungsprojekt (Special Research Project) "Synchronization 
of Civilizations in the Eastern Mediterranean in the 2nd millennium 
BC"; another conference was held at Haindorf in May 2001. 

Astrom chose the title "High, Middle or Low" for his symposium 
and thus placed his finger on the central question, which is still not 
solved today: the choice between a longer, shorter, or medium position 
for Hammurabi of Babylon. The conference in Gothenburg did close 
with a formal vote on which of the three alternatives the participants 
preferred: against 3 votes and 3 abstentions, the "Low" chronology was 
adopted, and it is absolutely clear for Egypt that for the NK, this is 
the only chronology with which we can live. There, I endeavoured to 
avoid the astronomical problems when discussing the chronology of the 
NK, and Kitchen also stresses in his most recent contribution that the 
Egyptian chronology "is not based on these meagre astronomical data". 30 
Helck was not a friend of astronomical data either. 31 

The apparent precision of astronomical and other sources from the 
natural sciences (including the ice of Greenland) is always enticing, and 
it is difficult to resist the charm. However, we should recall just how 
long it was generally agreed that the original introduction of the Egyptian 

28 Cf. Astrom, e<±, High, Middle or Low? 

29 Published in A&L 3 (1993). 

30 K. A. Kitchen, in: M. Bietak, ed., SCIEM 1996/98, 39. 

31 He stressed in "Zur Chronologiediskussion iiber das Neue Reich", A&L 3 (1992), 
63-67, that the dated monuments should be awarded priority before the astronomical 


calendar produced Eduard Meyer's "oldest date in world history" of 
19/VII/4241 BC — until this was demolished by Otto Neugebauer and 
Alexander Scharff in 1939. Even v. Beckerath's "earliest absolute date 
in Egyptian history" (Chronologie, 45) is an astronomical date (the heli- 
acal rising of Sothis at Illahun) and cannot be maintained. The con- 
troversy surrounding the Venus-dates in Mesopotamia, and the constantly 
renewed and alternatively calculated eclipse-dates in Western Asia (which 
do not touch Egypt) or the Sothis and lunar dates in Egypt have repeat- 
edly demonstrated the problems of the astronomical dates and con- 
tributed to the primacy of the purely historical dates. They are and 
remain our most important source. 

In an inscription in Karnak, the HP Osorkon proclaims that the cult 
was regularly performed, "even as the moon in its course" due to his 
efforts, 32 and thus the moon and its regular phases would appear to be 
the very pattern for precision. But lunar dates are repeated at quite 
short intervals and are hardly useful if they cannot be embedded in a 
fine mesh of other reference points, as is, for example, the case for the 
Illahun dates. From 21 lunar dates in that archive, R. Krauss has now 
calculated the alternative dates for the first year of Senwosret III at 
1862/61 or 1837/36 BC, and here it is the large number of dates 
which really brings weight since a single lunar date can bring no more 
than one further confirmation for a date which has been calculated 
using other means. Sothic dates with all of their countless parameters 
and uncertainties are likewise only of value as confirmations, and can 
never serve as the point of departure. And it almost never goes with- 
out any editing. 

In addition, the Egyptian scribes probably never aimed for such pre- 
cision; rather in their administrative tasks they will have been satisfied 
with rough approximations. It is thus that J. J. Janssen, who is pro- 
foundly familiar with their work, notes the "accuracy of the Egyptian 
scribes which is notoriously unreliable". 33 This will have applied to their 
methods of measuring time, which was extremely successful despite all 
its unreliability. The Egyptian calendar never depended upon interca- 
lations and is of winning simplicity. This applies not only to the con- 
stant length of the year, but also the schematic division of the year 
(three seasons, best given their Egyptian names, Akhet, Peret and Shemu, 

'fs III, pi. 16; Caminos, Chronicle, § 32 and 37f. 
Janssen, Varia, 101. 


each of four months of 30 days each). In Pharaonic times there was 
never an era with a constant continuous numbering of years: with each 
new Pharaoh, the count began anew. 

Egyptian scribes thus had no inhibitions about copying older texts 
in extenso without "modernizing" them, and the campaigns of a Pharaoh 
could be used by another later pharaoh; royal constructions could be 
newly dedicated with the addition of new names. But the framework 
of Egyptian history can only be relative, using contemporary dates from 
three millennia. Nevertheless here, encouraging progress has been made 
in the primary task of establishing a correct sequence of kings. 

As a festival celebrated (with the object of replenishing his powers) 
in the 30th year of a king's reign, 54 the ^-festival is chronologically 
relevant, and has frequently been included in chronological discussions. 
The apparent exceptions to the 30-year rule led to the assumption that 
this festival was regularly celebrated every 30 years, regardless of actual 
reigns. However, most of the exceptions have been eliminated, and 
there remain only a few uncertain cases (Hatshepsut, Akhenaten), which 
can be explained in terms of particular circumstances. In the case of 
Hatshepsut, the celebration can be linked to the accession of Thutmosis 
I (or II); in Akhenaten's case it can be anchored to the transition to 
the worship of Aten as king. A rigid 30-year sequence — as has occa- 
sionally been proposed — is improbable in the extreme. 

The most certain and best documented cases of a sed festival which 
was actually celebrated concern reigns where the chronology does not 
offer any difficulties (Amenhotep III, Ramesses II, Ramesses III). In a 
number of rather dubious cases, the alleged evidence of a ^-festival 
has been used to argue a reign of more than 3 decades, even where 
there is otherwise no evidence. This concerns Amenhotep II and 
Thutmose IV in particular, but all of the "sources" for the ^-festivals 
of these rulers are mere wishful thinking, although Wente and Van 
Siclen refer to "jubilee evidence in favor of a long reign for Thutmose 
IV". 35 They also take the alleged ^-festival of Hatshepsut as the basis 
for proposing a reign of 13 years for Thutmosis II. 36 A ^-festival is 
also proposed with regard to the length of the reign of Senwosret III. 3 ' 

On the issue of the date, etc., cf. Hornung & Staehelin, 

35 Wente & van Siclen, "Chronology", 230. 

36 Wente & van Siclen, "Chronology", 226. 

37 J. W. Wegner, "The Nature and Chronology of the Senwosret III-Amenemhat 
III Regnal Succession: Some Considerations Based on New Evidence from the Mortuary 


Such conclusions are too risky, and this also applies to the refer- 
ences to a "first time of the ^-festival" by kings of the MK and NK. 
Were we to use this in, e.g., the case of Amenhotep I as a reference 
demonstrating a reign of more than 3 decades, then we would come 
into conflict with the documented reign of 21 years, 18 and in the case 
of Psammetichus II whose reign is certain to have lasted 6 years, the 
"first time" cannot have any real meaning, and the same is true of 
Shoshenq I, likewise with 21 regnal years. 

With Hatshepsut, the "repetition" of sed- festivals, promised by Amun 
and Thoth in their divine speeches in the chapelle rouge at Karnak cannot 
be taken at face value. With Amenhotep II and Thutmose IV, "the first 
time and repetition of the ^-festival" is merely the expression of a 
wish, which cannot allow any historical conclusions. Only with a "sec- 
ond time of the ^-festival", known for Pepy II as well as the Archaic 
kings Den and Qa- C a, 39 do the sources depart from the wishful "repe- 
tition", and this also applies to the "3rd ^-festival" of Thutmose III 
and Amenhotep III, as well as the multiple repetitions documented for 
Ramesses II. 

The abundance of .^-festival wishes, documented for virtually all the 
dynasties, has no chronological importance. The .^-festival was merely 
understood increasingly as an ideal measure of time and used in par- 
allel with other concepts of time, precisely in order to wish Pharaoh 
the longest possible regnal era, whereby the hoped for chain of jubilee- 
festivals can be projected into the Beyond. Eduard Meyer already sug- 
gested "that this festival cannot be used for chronological purposes", 40 
and recent research gives us no reason to change this appraisal. 

Aside from the astronomical dates, the physical sciences offer a num- 
ber of other aids, among which C-14 radiocarbon dating has been 

Temple of Senwosret III at Abydos", JNES 55 (1996), 249-279, esp. 262-264, plead- 
ing for a real celebration. 

38 In the inscription of the "astronomer" Amenemhet, L. Borchardt, Die altdgyptische 
Zeitmessung (Berlin & Leipzig, 1920), PI. 18. E. F. Wente, "Thutmose Ill's Accession 
and the Beginning of the New Kingdom", JNES (1975), 265-272, esp. 271-272 sought 
to use the references to a .serf-festival and an alleged coregency with Ahmose to increase 
the length of the reign. 

39 For the references, cf. Hornung & Staehelin, Sedfest. For a new reference for Den 
on a bowl from Abydos, see G. Dreyer, "Umm el-Qaab: Nachuntersuchungen im 
fruhzeitlichen Konigsfriedhof 3./4. Vorbericht," MDAIK 46 (1990), 53-90, esp. 80, with 
Fig. 9. Against a 2nd .serf-festival for c Adj-ib (so A. Grimm "Ein zweites Sedfest des 
Konigs Acljib," VA 1 [1985], 91—98), cf. Kamil O. Kuraszkiewicz, "Noch einmal zum 
zweiten Sedfest des Adjib," CM 167 (1998), 73-75. 

40 E. Meyer, Geschichte des Altertums II, 1 (1953), 149, note 2. 


widely used, and become indispensable for Prehistory and the Archaic 
period. For the Dynastic period this procedure is, however, neither 
sufficiently reliable nor sufficiently precise. Climatic or seasonal dates 
such as harvests, the inundation, or preferred dates for expeditions in 
the desert regions from either bank of the Nile can provide useful 
checks, but we still lack a systematic and modern collection of the avail- 
able dates. Even today, historical periods and events continue to be 
explained in terms of climatic change, 41 although a causal relationship 
has never been demonstrated; the links with volcanic eruptions, such as 
Thera, have triggered debates, but not led to any definitive conclusions. 

Highly problematic are dates based on stylistic, linguistic and palaeo- 
graphic criteria. Here there are amusing discoveries, such as the dat- 
ing of a Hellenistic bronze of Socrates to "ca. 700 BC", which I noted 
in the Manchester Museum. The "Memphite Theology" has been 
pushed back and forth across the entire history of ancient Egypt from 
the Archaic period to the Ptolemies, like the unstoppable efforts to date 
the NK Books of the Netherworld to the MK or even the OK. In 
sculpture, there are still difficulties in the attribution of statues dating 
to the MK and NK and the LP. The dating of the handwriting of the 
Brooklyn oracle papyrus 47.218.3 can serve as an exemplary warning. 
In this case, 50 higher officials and priests all personally signed a doc- 
ument as witnesses, in Thebes on October 4, 651 BC (Julian). 42 Examined 
individually, one would date the signatures quite differently, but the 
date of the protocol clearly assigns them all to exactly the same single 

Synchronisms with the Near East are particularly useful when they 
can be linked to the relatively reliable Assyrian chronology. By con- 
trast, the numerous synchronizations with the Hittites are virtually use- 
less as the Hittite sources cannot provide either dates or regnal lengths. 
Our reliable point of departure remains the beginning of the reign of 

41 So, e.g., Th. de Putter, "Les inscriptions de Semna et Koumma (Nubie): niveaux 
de crues exceptionnelles ou d'un lac de retenue artificiel du Moyen Empire?", SAK 20 
(1993), 255-288 associates the frequently discussed extremely high Niles dating to the 
end of Dyn. 12 with an "evenement climatique"; for the movements of the Peoples 
of the Sea, cf. J. Neumann & S. Parpola, "Climatic Change and the Eleventh-Tenth 
Century Eclipse of Assyria and Babylonia", JNES 46 (1987), 161-182; a drought leads 
to famine and thus provokes the migrations. Cf. also S. J. Seidlmayer, Historische und 
moderne Nilstdnde (Berlin, 2001). 

*- R. Parker, A Saite Oracle Papyrus from Thebes in the Brooklyn Museum (Providence, 


Taharqa in 690 BC 43 As Depuydt shows in his contribution, the sequence 
with a "day-exact chronology" begins on June 20, 688 BC (Julian), 
with the sale of a slave. The recently discovered inscription of Sargon 
II in the Tang-i Var Pass in western Iran from the year 706 does not 
offer absolute precision about his immediate predecessors, as had been 
initially hoped, and thus alternatives remain (Shebitku as coregent or 
sole ruler). And there remain many uncertainties in the TIP, as critics 
such as David Rohl have rightly maintained; even our basic premise 
of 925 for Shoshenq's campaign to Jerusalem is not built on solid foun- 
dations. Nevertheless, there is such a web of dates, genealogies and 
relations between Egypt and the Near East that dramatic changes can 
be excluded (whereas Rohl wanted to cut off 141 years), above all due 
to the adjoining Ramesside era. In addition, there is archaeological 
material (such as, e.g., coffins) which has generally not been exploited 
for dating purposes, and there remains the prosopography of the officials 
and priests. For the TIP, there remain also the aids offered by the 
apis-bulls with their very precise data. 

Already at Gothenburg, there was general agreement about the dates 
for beginnings of the NK. Helck, Kitchen and Hornung/Krauss all 
worked with the very narrow range of 1540 to 1530 BC for the start 
of the reign of Ahmose, and after some debate, there is now general 
acceptance for the reign of Ramesses II at 1279—1213 BC. Although 
we must remain wary of confusing consensus with actual fact, for the 
NK we now have such a fine mesh of relative dates which are them- 
selves also woven into NE dates that major adjustments can probably 
be excluded. While there is room for minor cosmetic corrections, we 
are relatively confident about the framework. And now NE material 
allows for the fall of Babylon to be set at ca. 1500 BC. 44 We can now 
trust that dendrochronology will provide greater precision — as the pre- 
cise dating of the shipwreck of Uluburun with the Nefertiti scarab pro- 
viding grounds for hope. 4 ' The links in both directions — backwards to 
Amarna and forwards to the Ramessides — mean that even for the 
ancient crux of the length of the reign of Haremhab, those favoring a 

43 Although L. Depuydt now also allows for the possibility of 691 BC, "Glosses to 
Jerome's Eusebios as a Source for Pharaonic History", CdE 76 (2001), 30-31, note 1. 

44 Gasche et al., Dating. They argue for 1499 BC, but allow alternatives for 1507 
and 1491 BC. 

45 P. I. Kuniholm, B. Kromer et al., Nature 381 (1996), 780-783. Construction of 
the ship: 1316 BC. 


shorter reign are so hard pressed for explanations that the controversy 
would appear to have been silenced. In chronological issues we can 
never take a single item and redate it while disregarding the larger 
context in which it is fixed. Herein lies the great value of Manfred 
Bietak's endeavor to synchronize the civilizations of the 2nd mill, by 
bundling and weaving them together. 

After the departure of Parker and his "astronomically certain" dates 
there was a certain aporia with regard to the chronology of the MK. 
However, we will now have to take leave of Parker's date for the start 
of Dyn. 12 (1991 BC) and agree on a date around the middle of the 
20th century (although ca. 36 years still separate v. Beckerath and 
Krauss, with Kitchen in the middle). This offers encouraging possibil- 
ities for the extension of the FIP, which has hitherto been cut short. 
It is painful to recognize that the Near Eastern synchronism of Neferhotep 
I will have to be dropped, 46 but a certain degree of scepticism is always 
required when dealing with synchronisms. 

We are treading on very thin ice in the 3rd mill., even though fan- 
tastically precise dates based upon astronomical orientations of the pyra- 
mids are still being published. 47 While the general orientation does offer 
a certain framework, this is complemented by relative chronology. It 
no longer seems necessary to raise the dates for the pyramid-builders 
of Dyn. 4, to provide more building time and thus we can retain the 
23—25 years of the king-lists. At the top we can begin with Dyn. for 
the start of the Dynastic period in the 4th mill., which is essential for 
the synchronisms with the Near East. 

The most recent large-scale summary of Egyptian chronology is Jurgen 
von Beckerath's Chronologie des pharaonischen Agypten of 1997 which pro- 
vides a balanced state-of-the-art picture of the foundations. In contrast 
to Beckerath, we strive to separate clearly relative and absolute chronol- 
ogy, and to provide more weight for the archaeological materials and 
the factors derived from the physical sciences. 

Chronology has always been an arena for radical hypotheses and dras- 
tic moves. In antiquity, Christian chronographers manipulated Manetho's 
dates in order to achieve a convergence with the Biblical chronology. 

4b Cf. C. Eder, Die dgyptischen Motive in der Glyptik des ostlichen Mittelmeerraumes zu Anfang 
des 2. Jts. v. Chr. (Louvain: OLA 71, 1996), 13; T. Schneider, Review of Eder, "Mote", 
ZDPV 114 (1998), 184. 

47 Cf. the critical remarks by E. Aubourg, "Determination de l'orientation de la 
pyramide de Radjedef", Gemma 49 (2001), 245-248. 


At that time, the relative chronology was artificially lengthened whereas 
the 20th century has been marked by efforts at radical shortening, by 
eliminating dynasties or placing them in parallel with others; Velikovsky 
aimed at an extremely bold "analysis of events" by means of which 
similar events or historical constellations were simply slotted together — 
in this fashion Hitler's Russian campaign can be viewed as being the 
same as Napoleon's, and thus interpreted as the same event 

We will always be exposed to such attempts, but they could only be 
taken seriously if not only the arbitrary dynasties and rulers, but also 
their context, could be displaced. Were one to discover that Ramesses 
II was really Necho II in disguise, and likewise Merneptah as Apries 
and Ramesses III as Nectanebo I, one would still have to demonstrate 
that in each and every case the two allegedly identical rulers were also 
surrounded by the same officials, and that the religious and artistic 
contexts were also entirely compatible. In the absence of such proofs, 
we can hardly be expected to "refute" such claims, or even to respond 
in any fashion. For Dyn. 20 we have such a fine mesh of dates, vir- 
tually day-by-day, that the entire complex can hardly be assigned to 
another dynasty as it would simply fail. Furthermore, the entire so- 
called "dark ages" ca. 1200-700 BC are so well documented with 
archaeological material from Egypt that there is not the slightest ground 
for justifying a reduction. It is thus neither arrogance nor ill-will that 
leads the academic community to neglect these efforts which frequently 
lead to irritation and distrust outside of professional circles (and are 
often undertaken with the encouragement of the media). These attempts 
usually require a rather lofty disrespect of the most elementary sources 
and facts and thus do not merit discussion. We will therefore avoid 
discussion of such issues in our handbook, restricting ourselves to those 
hypotheses and discussions which are based on the sources. 

It is characteristic that the drastic reductions touching upon the Old 
Testament play a very important role, as in antiquity. In order to make 
Hatshepsut the Queen of Sheba and thus the contemporary of Solomon, 
we are obliged to remove the 500 years which separate these two by 
aligning a number of dynasties in parallel, rather than sequentially. 
Here we face those ideological pressures which always have an unsuit- 
ably powerful influence on the study of historical data. 

We have referred to the typically Egyptian association of the highest 
conceivable fullness of time with the solar orbit. "The lifetime of Re 
in heaven" was the absolutely highest conceivable temporal horizon for 
the ancient Egyptians, as the duration of being, comparable with the 


lifetimes of galaxies in modern astronomy. According to the Litany of 
Re (13th appellation) it was "greater than the West and its images", 
and thus actually greater than the duration of the Beyond. In the Book 
of the Celestial Cow, this concept is transformed into an image where 
neheh and djet "those two old and great gods" appear as the pillars of 
heaven. As long as time endures, heaven will rest on its pillars and the 
solar orbit will be maintained daily. Long before the Pythagoreans, the 
Egyptians had postulated the concept of the eternal return, which was 
finally given its impressive form by Nietzsche who was in turn draw- 
ing on Goethe: "How often 'tis repeated! will always be/repeated through 
eternity" (Faustus verses 7012f). But in Egypt, it was never the return 
of the same, but rather return in a form transformed: the sun is every 
day a new sun, reborn each day again by the goddess of heaven. 




The Editors 

Royal Annals of Memphis 

There are several fragments which almost certainly derive from more 
than one stone slab, inscribed on both sides. 1 The fragment with most 
text preserved is the so-called Palermo stone; other pieces are in Cairo 
and London. Nearly all of them were known and had been studied in 
the first decades of the last century, and some specialists have proposed 
reconstructions of the original text. 2 The inscriptions which record events 
in the reigns of Dyn. 1 through part of Dyn. 5 are arranged in a series 
of rectangular compartments, set out in horizontal rows, reading from 
right to left. Each compartment in row 1 contains only the name of a 
mythical or early historical king. In the other rows, the hieroglyph for 
"year" defines the right side of a compartment. Below the compart- 
ments, the height of the annual Nile inundation for that year is recorded. 3 

row 1 

Fig. I. 1.1. Rows 1 and 2 on the Palermo stone (Verso) 

1 Wilkinson, Annals, 28. 

2 Wilkinson, Annals, 29-36. 

3 Wilkinson, Annals, 18-19. 


The inscription of the first incomplete compartment of row 2 cites 
"following of Horus" and "Birth of Anubis". The second and third 
compartments are separated by a dividing line that indicates a change 
of reign. The second compartment lists "6 months 7 days", presum- 
ably the time elapsed in the last incomplete year of that reign. The 
third compartment lists "4 months 13 days", evidently with reference 
to the new king, but the two intervals do not add up to a full year be 
it lunar or solar. The compartment also records two rituals associated 
with accession years, "the Unification of the Two Lands" and the 
"Circuit of the Wall". 

For nearly all of the first three dynasties, the annals cite only occa- 
sions that gave a year its name. Beginning with Dyn. 4, the annals 
begin to resemble chronicles recounting a multiplicity of memorable 
events in each year of a reign. 4 The similarity of some so-called year 
labels of Dyn. 1 to certain entries in the annals suggests that both 
derive from the same archival institution.' Specialists do not agree on 
when the annals were compiled nor when the slab(s) were inscribed. 
Palaeography and the rendering of the royal names of the Early Dynastic 
Period favor of an OK date, but it cannot be excluded that the exist- 
ing fragments come from later copies. 6 

As Wilkinson observed, most scholars who have studied the annals 
have interpreted the information recorded at face value, instead of con- 
sidering the cultural context.' But even if cited events did not actually 
occur, the year itself may be accepted as fact. Being a later compila- 
tion, the annals are a less valid source for reconstructing regnal years 
of Early Dynastic kings than contemporaneous data would have been." 
A definitive reconstruction of the Annals as a whole is not possible, 
and their value for the Early Dynastic period is problematic. A revi- 
sion of the most recent attempt at reconstruction is made further below. 

South Saqqara Stone and Later Annals 

Annals of Dyn. 6 were inscribed on both sides of a slab measuring ca. 
226 X 92 cms that was reused as a sarcophagus lid. At that time the 

Clagett, Science I, 48. 
Wilkinson, Annals, 60. 
Wilkinson, Annals, 23-24. 
Wilkinson, Annals, 65. 
Spalinger, "Texts", 281. 


text was summarily erased. Baud and Dobrev analyzed the structure 
of the text and suggested readings of details. 9 Traces can be identified 
of the titularies of the Dyn. 6 kings Teti, Userkare c , Pepy I, and Merenre c 
which show that the count was biennial during their reigns. 

From later times there is a portion of the annals of Amenemhet II 
containing chronicle-like entries. 10 The so-called annals of Thutmose III 11 
report on a daily basis, whereas the others provide yearly overviews. 12 
The annals of Pami list offerings donated to the cult in Heliopolis dur- 
ing his reign. 13 

Proposed Revision of Recent Reconstructions of the Memphite Annals Stone 
(Fig. I. 1.2) 

The latest reconstruction of the annals proposed by Beckerath, 14 can 
be improved by taking into account Baud's corrections. 1 ' Any attempt 
at reconstruction must be based on a sound order of succession. This 
is the case for the kings of Dyn. 1 and for Hetep-sekhemwy, Ra c -neb, 
and Ny-netjer of earlier Dyn. 2, since Kahl established that Weneg is 
the nsw bjt name of Ra c -neb. 16 The succession in the second half 
of Dyn. 2 is unclear. Traces of a serekh with the Seth animal on Cairo 
1 suggest that Per-ibsen followed Ny-netjer, but it is also possible that 
Egypt was divided at that time. If so, the Annals might list the kings 
in succession although they actually ruled simultaneously. It has recently 
been confirmed that Kha-sekhemwy's successor was Netjery-khet (Djoser), 
not Nebka. 1 ' The sequence Kha-sekhemwy: Djoser is recorded in row 5 
of the Palermo stone. Djoser's successor Sekhem-khet may be identifiable 

9 M. Baud & V. Dobrev, "De nouvelles annales de l'Ancien Empire egyptien. Une 
<Pierre de Palerme> pour la VT dynastie", BIFAO 95 (1995), 23-92; idem, "Le Verso 
des annales de la VF dynastie. Pierre de Saqqara-Sud", BIFAO 97 (1997), 35-42; see 
also below Baud, Chapter II. 5. 

10 H. Altenmuller & A. M. Moussa, "Die Inschrift Amenemhets II. aus dem Ptah- 
Tempel von Memphis", SAK 18 (1991), 1-48.— J. Malek & S. Quirke, JEA 78 (1992), 

11 Urk. IV 645-673. 

12 W. Helck, LA I, 279. 

13 S. Bickel, M. Gabolde & P. Tallet, "Des annales heliopolitaines de la Troisieme 
Periode Intermediate", BIFAO 98 (1998), 31-56. 

14 Beckerath, Chronologie, 174-179. 

15 Baud, "Menes", 136-138. 

16 See below Kahl, Chapter II. 2. 

17 See below Seidlmayer, Chapter II. 3. 



on the large Cairo fragment; the succession at the end of the dynasty 
is unclear. 

Maximum and minimum distances between the Palermo stone and 
Cairo 1 can be determined on the basis of the preserved compartments 
for [Djoser] in row 5. Beckerath's change of reign after the first com- 
pletely preserved compartment on Cairo 1 is spurious. 1 " Rather, there 
are Va + 9 compartments on Cairo 1 that belong to [Djoser]. If the 
lost titulary was centred and had a width of 7 compartments, then at 
least '/a + 9 + 7 + Va + 9 = 26 compartments result for Djoser. If 
the first of the 26 compartments lay to the left of the Palermo stone, 
then the gap is 17 compartments or more. If the first of these com- 
partments was identical with the first of [Djoser's] preserved '/a + 4 
compartments on the Palermo stone, then the gap measures at least 
12 compartments in row 5, exceeding the 9 compartments deduced by 
Wilkinson. 19 A gap of 12 compartments in row 5 is not compatible 
with a gap of 9 compartments in row 2. According to Kaiser the inter- 
nal evidence of the Palermo stone and Cairo 1, including the corre- 
spondence of rows 2 to 5, favours 9 missing compartments in row 2 
as does the regularity of the "followings of Horus" and the six-yearly 
bark festival in rows 2 to 4. a) We accept Kaiser's arguments for the 

Fig. I. 1.2. Revised reconstruction of the Annals stone (Verso) 

18 Wilkinson, Annals, 53. 

19 Wilkinson, Annals, 79. 

20 W. Kaiser, %AS 86 (1961) 



size of the gap in row 2 and the corresponding gap of 13 compart- 
ments in row 5 in Beckerath's reconstruction. 

The left edge of row 2, and thus of the other rows on the recto (if 
row 3 begins with "Serpent") can be determined with reference to the 
completely preserved titulary of Djer that spans 7 compartments. Taking 
into account the gap of 9 compartments in row 2 between Cairo 1 
and the Palermo stone, there were 20 compartments before and after 
Djer's centred titulary, for a total of 47 complete compartments. 

The determination of the right edge of the inscribed field is more 
difficult. Beckerath errs in identifying Kha-sekhemwy on the Palermo 
stone as Nebka and in construing the right edge of row 5 by adding 
the TC's 27 years of Bebti (<Kha-sekhemwy) to the right of **Nebka's 
compartments. By contrast, extrema for the right edge can be deter- 
mined in row 6 by extrapolating the early years of Snofru. The min- 
imum number of compartments is 7 {zp years 1 to 6 and a year zmj-tHwy), 
and the maximum 12 (adding m-jit zp years 1 to 5). Thus at least 16 
and at most ca. 27 compartments belonged in row 2 to the right of 
Djer's reign. We opt for the maximum 75 compartments, correspond- 
ing to 74 years, because the incomplete last year of [ c Aha] and Djer's 
first one occupy two compartments. The recent discovery of a year 
tablet of Nar-mer makes it possible that row 1 recorded not only c Aha, 
but also Nar-mer. 21 

The reign of Semer-khet is completely preserved in Va + 1 + Va com- 
partments in row 3 on Cairo 1. Towards the left, ca. 24 compartments 
can be reconstructed for Qa- C a as the successor of Semer-khet, if the 
47th compartment for Djer marks the left edge of the inscribed field. 
To the right of Semer-khet, half of a compartment and part of the last 
complete compartment of c Adj-ib are preserved without traces of his 
titulary; Wilkinson postulates at least Va +7 + Va compartments for him. 

Beckerath assumes that Den celebrated the Sed festival which is 
attested on the Palermo stone in his 30th year. The assumption is unac- 
ceptable, if only, because Qa- C a celebrated two Sed festivals according 
to contemporaneous sources, whereas he ruled less than 30 years accord- 
ing to Beckerath's own reconstruction. The end of Den's titulary is pre- 
served on the Palermo stone and its beginning on Cairo 5. 22 If Den's 
titulary had a width of 7 compartments, then at least Va + 13 + 7 + 

21 Wilkinson, Annals, 72. 

22 Note, however, that Cairo 5 may be a forgery. 


13 + V2 — 34 years are deducible, resulting in V2 + 1 1 + V2 years for 
c Adj-ib. At most V2 + 17 + 7 + 17 + V2 compartments can be pro- 
posed for Den, if c Adj-ib reigned a minimum of 8 years. 

If the right edge of the inscribed field is defined by 1 2 reconstructed 
compartments for Snofru in row 6, then to the right of Den there 
should be at least 8 and at most 12 compartments for "Serpent". The 
reconstruction results in at most ca. 90 compartments in row 3. 

Row 4 preserves a series of counts, combined with "followings of 
Horus" from the reign of Ny-netjer whose titulary is partially preserved. 
The example of Djoser in row 5 shows that the first count could occur 
as late as the second year. If so, then ca. Va + 16 + 7 + 16 + V2 — 
40 compartments are possible. If the width of the centred titulary 
amounted to 8 compartments, then 41 compartments result. The recon- 
struction yields a gap of 2 compartments at most between the 41st 
compartment of Ny-netjer and the following reign on Cairo 1. The 
dilemma cannot be solved by making the distance between the Palermo 
stone and Cairo 1 smaller, because we already presume the minimum 

The position of the titulary of Per-ibsen (?) as Ny-netjer's successor 
implies that the former's reign spanned about 10 compartments. The 
remainder of row 4 towards the left edge amounts to ca. 22 compart- 
ments (with the width of the Cairo 1 compartments), to be distributed 
among the kings of Dyn. 2. 

Between Ny-netjer's reconstructed first year and the right edge of 
row 4 as defined by Snofru's reconstructed first compartment, there 
remain ca. 29 compartments (width of Palermo stone compartments) 
for Ra c -neb and Hetep-sekhemwy. A reconstruction along these lines 
results in a total of at least ca. 104 compartments in row 4. 

In row 5 the reconstruction of the preserved counts that are com- 
bined with "followings of Horus", allows the reconstruction of at least 
10 lost compartments of [Kha-sekhemwy] and of another 1 + '/a, if 
the first count and "following of Horus" occurred in the second reg- 
nal year. The result is 18 compartments for [Kha-sekhemwy], leaving 
ca. 1 7 compartments towards the right edge of row 5 for kings of late 
Dyn. 2. For [Djoser] V2 + 26 + V2 compartments can be reconstructed 
and 7 compartments for [Sekhem-khet] as his successor. The remain- 
ing ca. 16 compartments at the left end of row 5 belong to the last 
three kings of Dyn. 3. The reconstruction yields altogether 86 or 87 
compartments at most. 


About 19 compartments should follow after Snofru's mpt zp 8 on the 
Palermo stone, if Djer's 47th compartment defines the left edge of the 
inscribed field. Reconstructed row 6 contains a maximum of ca. 34 
compartments for Snofru. 

Rows 7 and 8 of the recto of the Annals are mostly destroyed. Cairo 
fragment 3 shows Djedefre' occupying the last third of row 8, whereas 
the rest of it and all of row 7 must have belonged to [Cheops]. The 
short reign of Bicheris might have been accommodated at the end of 
row 8. If rows 7 and 8 amounted to more than 30 years for Cheops 
and Djedefre c , then row *9 would not be long enough for Khephren, 
and Beckerath is right in postulating a row *10. (Row 1 would yield 
at least 100, and at most 123 compartments, possibly corresponding to 
the 115 mythical kings listed in the TC before Menes.) 

Altogether we estimate ca. 75 compartments (= 74 years) in row 2, 
ca. 90 compartments in row 3, ca. 104 compartments in row 4, and 
ca. 87 compartments in row 5, corresponding to ca. 164 years for Dyn. 
1, ca. 141 years for Dyn. 2 (if the reigns recorded were successive), 
and ca. 50 years for Dyn. 3, or a total of ca. 355 years for Dyns. 1 
to 3. The result is virtually the same as Kaiser's figure of 359 years, 
but markedly less than Beckerath's 393 years. The reliability of any 
sum is, however, open to question; whereas the order of succession as 
preserved in the Annals conforms to contemporaneous data, its rele- 
vance for counting regnal years remains conjectural. 


Mm Ryholt 


The King-list or "Royal Canon of Turin" is the only true king-list pre- 
served from ancient Egypt before the Ptolemaic period. It is a "true 
king-list" in the sense that the compiler of the document aimed at 
recording all of the kings of Egypt along with their reign-lengths. This 
stands in striking contrast to the other lists, such as the cultic assem- 
blages of deceased kings engraved on walls in the temples of Sety I 
and Ramesses II at Abydos and in the tomb of the priest Tjuloy at 
Saqqara, 2 even if these lists preserve the royal names in forms that are 
superior to those of the more or less contemporaneous TC 

Only two studies on the nature of the TC as such have been pub- 
lished; the first by Redford in 1986, 3 and the other by myself in 1997. 4 
Studies of the chronological implications and attempts at reconstruc- 
tions of the list have been more numerous. The most significant advances 
in the reconstruction touch the Late OK, FIP and the SIP.' 

Around 1820 Bernardino Drovetti, French Consul in Egypt, acquired 
the TC, which eventually passed to the Turin Museum. The script is 
hieratic, written on the reverse of a discarded tax register dating to the 
reign of Ramesses II. Thus the king-list could have been written at the 
earliest in the time of Ramesses II himself. The TC can be accurately 

1 For detailed discussion of the Turin king-list, see K. Ryholt, A&L 14 (2004), 

2 See Redford, Annals, 18-24. 

3 Redford, Annals, 1-18. 

4 Ryholt, Situation, 9-33. 

5 Late OK: K. Ryholt, Z^S 127 (2000), 87-100. FIP: J.v. Beckerath, £4'S 93 (1966), 
18-19. SIP: Ryholt, Situation, 69-75, 94-97, 118-119, 151-159, 163-165; for Frg. 134, 
cf. J. P. Allen, BASOR 315 (1999), 49-50.— W. Helck, SAK 19 (1992), 150-216, pro- 
posed a series of new arrangements, none of them based on autopsy. I could verify 
none of them whereas several can definitely be rejected, cf. Ryholt, Situation, 2 1 . Note 
that Beckerath, Chronologic, 20, follows Helck's reconstruction. 



described as a copy, drawn up in a somewhat careless manner on waste 
paper, from a damaged and imperfect original. The scribe apparently 
did not attempt to supplement missing information from other sources 
which may have been available at the time. 6 Apart from a section cut 
off in antiquity, the papyrus was presumably intact upon discovery. It 
has since been reduced to more than 300 fragments, apparently due 
to rough handling. Since Champollion hrst saw and described it in 
1824, the papyrus has deteriorated considerably. However, thanks to 
the efforts of several scholars over the past 150 years, most of the larger 
fragments have now been joined. Yet, many of the very small scraps 
remain unpublished, and the position of many important fragments 
remains conjectural. 

The papyrus has a large format, measuring 42 cm in height and 
about 1 3 A m in length. There are now 11 columns; one or two were 
lost in antiquity when the piece of papyrus was cut off. It could have 
included Dyns. 17, 18 and part of Dyn. 19. What follows refers to a 
new reconstruction, in progress (Table I. 2.1). 7 

Table I. 2.1. Concordance between Gardiner's edition 
and the new reconstruction 

Column 1 
Column 2 

Column 3 
Column 4 
Column 5 
Column 6 
Column 7 
Column 8 
Column 9 
Column 10 

Column 1 1 

Gardiner col. I 

Includes Frg. 41-42 (Gardiner col. IX) and Frg. 150-152 
and Frg. 22+unnumbered fragment (Gardiner col. X) 
Gardiner col. II 

Gardiner col. 


Gardiner col. 


Gardiner col. 


Gardiner col. 


Gardiner col. 


Gardiner col. 


Includes Frg. 105+108 (Gardiner, col. IX), Frg. 138 
(Gardiner, col. X) and unnumbered fragments (Gardiner, 
col. X. 13-21) 
Gardiner col. XI 

6 Thus, for instance, the names of ten late OK kings are lost from TC and the 
damaged writings of the names of two following kings are intact in the Abydos List. 

7 For the new arrangement of the SIP section (columns 7-11) see Ryholt, Situation, 
69-75, 94-97, 118-119, 151-159, 163-165. 



Groupings of Kings; Headings and Totals 

The TC lists three main categories of "rulers": 

(1) god and demigod c. IV2 columns Cols. 1-2 (bottom) 
kings (ntr.w): 

(2) spirit kings (>h.w): c. V2 column Cols. 2 (bottom)-3 

(top half) 

(3) historical or human 8V2 columns Cols. 3 (bottom half)-ll 
kings (~rmt.w): 8 

Each ruler's name and the length of reign, either precisely in years, 
months and days, or in years alone, was recorded. 9 The variations in 
detail provided for individual kings indicates that the Vorlage made use 
of a number of sources with different formats (Table I. 2.2). The most 
information is provided for the archaic kings; their reigns were recorded 
in years, months and days and their ages at death were also noted. But 
the royal names in this section are the most severely corrupted. The 
details are therefore not necessarily indicative of superior source material. 

Table I. 2.2. Division of the Turin King-list into sections 



Details provided about kings 


Archaic Period 

Dyns. 1-2 

Years, months and days + Age at death 



Dyns. 3-6 

Years alone 10 


Late OK & 

Dyns. 7-8 

Years, months and days 

FIP (Herakleopolis) 

Dyns. 9-10 


FIP (Thebes) 

Dyn. 11 

Years alone 


MK & 

Dyn. 12 

Years, months and days 


Dyns. 13-16 

Explicit information about the nature of a given group of kings is pro- 
vided by headings, most damaged. The coherent bits of what remains 
may be translated as follows: 11 

8 This division of the kings is based on Manetho's terminology, cf. Waddel, Manetho, 

9 By contrast to Manetho gender goes unremarked. The TC included at least one 
female ruler, viz. Nofrusobk (7.2). Nitocris of Dyn. 8, who has hitherto been consid- 
ered a woman on the basis of later tradition, seems to have been male; see K. Ryholt, 
ZAS 127 (2000), 92-93, 99-100. 

10 There are two anomalies within this section where both months and days are 
recorded as well, i.e. TC 4.7 (a difficult royal name, or perhaps rather the record of 
a lacuna, recorded between Sekhem-khet and Huni) and TC 5.1 (Teti). 

11 Redford, Annals, 11-13, and Helck, SAK 19 (1992), 151-216, offer different recon- 


Table I. 2.3 

Heading for Dyns. 1-5 (3.10) 

"[ICings of the house of (?)] King Menes." 

Totals for Dyns. 1-5 (4.26) 

"Total of kings from Menes until [Wenis: x amounting to 767 (or 768) years.]" 12 

Totals for Dyns. 6-8 (5.14-15) 

'[Total of] kings [until Neferirkare': x] amounting to 181 years, 6 months, 
3 days, and a lacuna of 6 (years). Total: 1 [87 years, 6 months, and 3 days]." 

Totals for Dyns. 1-8 (5.15-17) 

"'[Total of] kings 13 [from] Menes; their kingship, their years, and a lacuna 
[thereto]: 9 [4] 9 years and 15 days, and a lacuna of 6 years. Total: [x kings 
amounting to] 955 years and 1 [5] days." 

Totals for Dyns. 9-10 (6.10) 
"Total: 18 kings . . ." — rest lost 

Heading for Dyn. 11 (6.11) 
"Kings of. . ." — rest lost 

Totals for Dyn. 11 (6.18) 

'[Total:] 6 kings who ruled 1 [36 years] and a lacuna of 7 (years). Total 143 

Heading for Dyn. 12 (6.19) 
"[Kings of] the residence 'It-t>voy." 

Totals for Dyn. 12 (7.3) 

"Total of kings of the residence ['It-t>wy\: 8 who ruled 213 years, 1 month 

and 17 days." 

Heading for Dyn. 13 (7.4) 

"Kings [who were] after the children (?) [of Dual] King [Sehet]epibreV 14 

Totals for Dyn. 15 (c. 10.29) 

"[Total:] 6 [Hyk]sos. They ruled 1[0]8 [years]." 

structions and translations. On the reconstruction of summations for Dyns. 6-8 and 
1-8, see W. Barta, MDAIK 35 (1979), 13-14, and K. Ryholt, Z^S 127 (2000), 91, 

12 There is not space enough for this figure to have included years, months and 
days; possibly even the years were excluded and simply the number of kings recorded. 
The number of years is based on the summations for Dyns. 6-8 and 1-8. 

13 Lit. "kingship"; clearly an error. 

14 Reading fhrj-s,' hrdf.w], see Ryholt, A&L 14 (2004), 142, n. 38. An alternative 
reading [hr]-s> msf.w], which produces the same sense, was proposed by J. P. Allen at 
the colloquium The Second Intermediate Period: Current research, future prospects at the British 
Museum, 2004. 


There is no heading for the kings of Dyn. 6, nor for the Herakleopolitan 
kings or those of Dyn. 14. Chronological details important from a 
modern perspective go unmentioned, above all, information on over- 
lapping dynasties and coregencies. When two groups of kings overlap, 
they are simply listed consecutively. For example, the first Theban kings 
of Dyn. 1 1 follow the last Herakleopolitan rulers with whom they were 

The reign-lengths of Dyns. 3—6 and 9—10 recorded in full years alone 
pose a special problem. It remains unclear how these figures were 
rounded off. The reliability of the totals provided for groups of kings 
is intimately related to the accuracy with which the individual reigns 
were recorded. The totals provided for Dyns. 6-8 and 1-8 are subject 
to the same factors. The former total again includes a lacuna and both 
include a series of kings whose reigns are expressed only in years while 
others have reigns recorded in years, months and days. The total for 
Dyns. 1—8 includes no less than 26 kings recorded by years alone, 
resulting in a maximum margin of error of a quarter century. The part 
which appears to have suffered the most is the Late OK section (Dyns. 
7-8). As noted above, a group of ten consecutive kings is entirely lost 
and the names of the next two are only partly preserved. 15 The lost 
kings are accounted for by the word wsf ("lacuna") in both the total 
for Dyns. 6—8 and that covering all of Dyns. 1—8. 

The total of Dyn. 1 1 is recorded in full years only, obviously because 
the kings reigns were so recorded. Since the fractions of individual 
reigns are omitted, the total is inevitably approximate, with the mar- 
gin of error amounting to few years. The error might be more significant 
if "Mentuhotep I", who never actually reigned as king, was assigned a 
fictitious reign-length. Finally, there is the lacuna referring to Mentuhotep 
IV and the question of the accuracy of the 7 years ascribed to him in 
the total for Dyn. 1 1 . 

There is a certain amount of evidence that several Dyn. 12 kings 
reigned jointly and that some coregencies could have lasted up to a 
decade. Yet the scribe who calculated the sums for the individual groups 
of kings was not aware of (or did not take account of) the overlap- 
ping reigns. He simply added the full lengths of individual reigns to 
calculate the total for the dynasty. The overlapping reigns probably 

15 J.v. Beckerath, JNES 21 (1962), 140-147; K. Ryholt, %AS 127 (2000), 96-99. 


accounted for a bit more than 2 decades. Furthermore the TC's infor- 
mation about Dyn. 12 is disquietingly incompatible with the ample 
contemporaneous sources. lb 

In the sections for Dyns. 13 and 14, the notations of at least two 
lacunae are preserved. One is recorded after the entry for Sonbef (7.6) 
and the other after Nebsenre' (9.14). It is not clear how many kings 
were lost, but at least one king can be identified by name in relation 
to the first lacuna, viz. Nerikare'. 17 Further kings may have been lost 
in the same sections. 1 " There are two instances where reign -lengths are 
partially damaged. One is ?w-ib-r c of Dyn. 14 whose reign is recorded 
as "[x years], lacuna, 18 days" (9.12) and the other is swsr-n-/ of Dyn. 
16 with "12 years, lacuna, [x] days" (11.8). 

Relation to Manetho's King-list 

It is significant that the five sections A— E in Table I. 2.2 all correspond 
to the dynastic arrangements Manetho adopted. Section A equates to 
Manetho's Dyns. 1-2, Section B to his Dyns. 3-6, Section C to his 
Dyns. 7—10, Section D to his Dyn. 11, and Section E to his Dyns. 
12—16. The correspondence becomes even more striking when the tex- 
tual division of the kings into groups is taken into account. The TC 
groups the following kings: Dyns. 1—5, Dyn. 6, Dyns. 7—8, and Dyns. 
9—10, followed by Dyns. 11, 12, 13, 14, 15 and 16. The main difference 
from Manetho's list is that the TC does not split up Dyns. 1-2 (Archaic 
Period), Dyns. 3—5 (Memphis), Dyns. 7—8 (Memphis), and Dyns. 9—10 
(Herakleopolis). 19 The few reigns of the OK that were not corrupted 
beyond recognition in Manetho seem to be either rounded up or down 
in comparison to the TC. Hence Manetho's figures seem to be based 
on a tradition with reigns recorded in years and months if not also in 

16 See below Schneider, Chapter II. 7. 
'' Ryholt, Situation, 318. 

18 Ryholt, Situation, 70 (Dyn. 13), 94—95 (Dyn. 14). Alternative interpretations have 
been presented by C. Bennett, GM 159 (1997), 11-17 and J. P. Allen, BASOR 315 
(1999), 50-51. 

19 Cf. below Seidlmeyer, Chapter II. 3. 



Despite its immense historical value, the TC is in various ways far 
removed from an ideal source. The incomplete state of preservation 
and inadequate publication inhibit full access both to the information 
it once contained and to what remains. There are also several intrin- 
sic features that detract from its value as chronological source. If the 
scribe's priorities were historical, he was clearly not primarily concerned 
with either absolute or relative chronology. This is obviously not an 
ideal point of departure, and it warns us that there may be errors that 
we are presently unable to verify lurking in the document. Accordingly, 
the text should be treated with circumspection. 


The Editors 

Comprehensive king-lists are engraved on walls in the temples of Sety 
I and Ramesses II at Abydos, at Karnak and in the tomb of the priest 
Tjuloy at Saqqara. 1 They list kings in an apparently correct sequence 
and render their names in forms superior to those of the more or less 
contemporaneous TC. In the Abydos list, the names of ten kings lost 
from the Late OK section of the TC and two damaged names that 
follow are recorded intact. 2 By contrast, the king-list at Karnak from 
the time of Thutmose III, though mentioning a number of rulers omit- 
ted in other lists, does not give the names in historical sequence. Since 
all these lists served the cult of the deceased kings, the names sufficed, 
and further information, such as regnal years, was not necessary. None 
mentions kings considered to be illegitimate, for example Hatshepsut 
or the Amarna rulers. The later king-lists of Greek historians seem to 
derive from the annalistic tradition, rather than from such cultic king- 

In Book II of his History, Herodotus reported the names and correct 
reign lengths for the kings of Dyn. 26, and he stated that there were 
330 earlier kings, including 18 Ethiopians and Queen Nitocris, but his 
information on the period prior to Dyn. 26 is virtually useless for recon- 
structing Egyptian history and chronology. 

A king-list with 38 + 53 names and regnal figures is attributed to 
Eratosthenes. 3 Apollodoros preserved 38 names that were copied by 
George the Monk, known as Syncellus. No. 29 is easily recognizable: 
XQMAEO0A < Sty mr.n Pth, i.e. Sety I who is given 1 1 regnal years, 
as is no. 36 ZIOOAZ < *ZIO0AX, i.e. Siptah with 5 regnal years. 4 The 
names are accompanied by their secondary Greek translations, added 
only after noticeable corruption of the text occurred. 

For the Abydos and Saqqara lists, see Redford, Annals, 18-24. 

Cf. above, Chapter I. 2. 

Waddell, Manetho, 213-225. 

Krauss, Amarnazeit, 274-276, with additional literature. 


Diodorus utilized Herodotus and other sources for the chapters on 
Egypt in his Bibliotheca Historical As a rule he did not list regnal years, 
except for Kheops and Khephren where he follows Herodotus. Diodorus 
mentions five ruling Egyptian queens instead of Herodotus's single 

Manetho's Aigyptiaka is purportedly the work of a priest who was a 
contemporary of Ptolemy II. 6 The Aigyptiaka displays remarkable simi- 
larity to the Turin King-list: rulers, including illegitimate kings recorded 
by name with reign length, arranged in groups, and listed in a sequence.' 
But there are two features which are specific to the Aigyptiaka and which 
have no counterpart in pharaonic tradition: glosses and antisemitism. 
The former reflect primarily Herodotus and Diodorus, 8 which suggests 
that the list may have been compiled after Diodorus. The antisemitism 
is manifest in the story of the enemies of king Amenophis, i.e. the lep- 
ers of Egyptian origin and the descendants of the Hyksos as common 
forefathers of the Jews. 9 Antisemitism is not attested before the Macca- 
beans; 10 therefore the story cannot be ascribed to Manetho in the 3rd 
century BC, but should rather be dated to the 1st century AD. The 
traditional explanation that Manetho made use not only of authentic 
records, but also of popular romances devoid of historical value, 11 does 
not take into account the motive of antisemitism and its history. 

It is a fact that the Aigyptiaka was cited by none of the great com- 
pilers of antiquity like Pliny, Diodor, and Strabo, nor was it used by 
any known Alexandrian scholar. 12 When Apion cited the destruction of 
Avaris by c Ahmose according to the Egyptian historian Ptolemy of 
Mendes, 13 he did so without mentioning the differing version in the 

5 E. Murphy, The antiquities of Egypt: a translation with notes of book I of the Library of 
history of Diodorus Siculus, rev. and expanded (New Brunswick, 1990). 

6 G. Verbrugghe & J. M. Wickersham, Berossos and Manetho (Ann Arbor, 1999), 

7 For Manetho's king-list, see Waddel, Manetho, and F. Jacoby, Die Fragmente der 
Griechischen Historiker, III C (Berlin & Leiden, 1958), no. 609. 

8 L. Stern, "Die Randbemerkungen zu dem manethonischen Kbnigscanon", £AS 
23 (1885), 87-96. 

9 M. Stern, Greek and Latin authors on Jews and Judaism I III (Jerusalem, 1974-1984), 

10 I. Heinemann, "Antisemitismus", in: Pauly-Wissowa, RE, Supplement vol. 5 (1931), 

11 Meyer, Chronologie, 78-79; Gardiner, Egypt, 47; Beckerath, Chronologie, 35. 

12 G. F. Unger, Chronologie des Manetho (Berlin, 1867), 3; 116. 

13 A. Dihle, "Ptolemaus von Mendes", in: Pauly-Wissowa, eds., RE, Reihe 1, Vol. 
23.2 (1959), 1861. 


Aigyptiaka. The first to cite the Aigyptiaka was Flavius Josephus in Contra 
Apionem; subsequently, the christian chronographers Africanus and 
Eusebius cited it; still later Syncellos contributed greatly to the trans- 
mission. 14 The silence of authors earlier than Josephus and the anachro- 
nisms in the text arouse suspicion that it is a pseudepigraphic work, 
based on the authentic tradition of the annalistic king-list, but com- 
piled after Diodorus. 15 

Gardiner's comments on the Aigyptiaca can be paraphrased as follows: 16 
In Manetho's work the entire history of Egypt, after the reigns of the 
gods and demi-gods, was divided into 31 dynasties of royal families, 
beginning with Menes and ending with Alexander the Great's conquest 
in 332 BC In spite of all the defects this division into dynasties exhibits, 
it has taken so firm a root in the literature of Egyptology that there 
is but little chance of its ever being abandoned. In the forms in which 
the book has reached us there are inaccuracies of the most glaring 
kind, these finding their climax in Dyn. 18, where the names and true 
sequence are now known from contemporary sources. Africanus and 
Eusebius often do not agree; for example Africanus assigns nine kings 
to Dyn. 22, while Eusebius has only three. Sometimes all that is vouch- 
safed to us is the number of kings in a dynasty (so in Dyns. 7-10, 20) 
and their city of origin. The royal names are apt to be so incredibly 
distorted, that of Senwosret I of Dyn. 12, for instance, being assimilated 
in the form of Sesonchosis to that of Shoshenk of a thousand years 
later. The lengths of reigns frequently differ in the versions of Africanus 
and Eusebius, as well as often showing wide departures from definitely 
ascertained figures. When textual and other critics have done their best 
or worst, the reconstructed Manetho remains full of imperfections. 

None the less, Manetho did preserve details which have been con- 
ceded as historical by Egyptologists if only very late. An example is 
the name of a king Nephercheres, whom Manetho had placed in Dyn. 
21; in the 1930s a small object bearing the name was found among 

14 See in detail Beckerath, J^wischenzeit, 1 1-20; idem, Chronologie, 35-38, assuming 
that the Aigyptiaka is authentic. 

15 Waddell, Manetho, VIII, XIV; E. Hengstenberg, Manetho und die Hyksos. Beilage zu: 
Die Biicher Moses und Agypten (Berlin, 1841); R. Krauss, "Manethos Agyptische Geschichte — 
eine ptolemaische oder romische Kompilation?", in: Timelines. Studies in honour of Manfred 
Bietak, E. Czerny et al., eds., (Leuven: OLA 149.3, 2006), 227-234. 

16 Gardiner, Egypt, 46-47. 


the grave goods of Psusennes I in Tanis." Another example is the rul- 
ing queen Akencheres (< c Ankhetkheprure c ) of Dyn. 18, the daughter 
of a king, whose historical existence was archaeologically first confirmed 
in the 1970s. 18 Manetho's reign lengths are at times correct to the 
month, for example in the case of Ramesses Miamun, i.e. Ramesses 
II, who reigned for 66 years and 2 months according to Josephus's 
copy of Manetho. On the other hand, Manetho's pre -Dyn. 18 regnal 
figures often seem to be systematically distorted by multiples of ten, 19 
probably in an attempt to bring Egyptian chronology into line with a 
version of biblical chronology. 

17 Cf. below, Chapter II. 9. 

18 Cf. below, Chapter II. 8. 

19 Helck, Manetho, 81-83. 


Morris L. Bierbrier 

Genealogy, the study of family pedigrees, can provide a useful fix on 
proposed chronological schemes for ancient Egypt. Obviously the pas- 
sage of time can be limited or, in certain instances, can be expanded, 
by the knowledge that an individual, his family, or a stated number of 
generations of a family, lived within a set period. It has been calcu- 
lated that a generation might comprise 25—35 years on average, but, 
in the surviving ancient Egyptian documentation, we are not neces- 
sarily dealing in averages. 1 Thus the possibility remains that in certain 
cases an elderly father might produce a son who in turn might live to 
a great age and so throw any calculations off. A complex and inter- 
linked genealogy might overcome this possibility but such genealogies 
are rare. 

The use of genealogies in chronological research is limited by sev- 
eral factors. The most important of these is the lack of documentation. 
This is especially true of those periods when the chronology is most 
uncertain. In other periods, abundant documentation can lead to con- 
fusion as the same names are used repeatedly and identification of 
different generations may be uncertain. The documentation normally 
only refers to the elite families although occasionally at such places as 
Kahun or Deir el-Medina, information on lower-ranking families is 
forthcoming. It is now clear that the Egyptians practised some form of 
ancestor worship and records were certainly kept by some families of 
their ancestral lines although such information was not necessarily 
recorded on stone and thus has not survived to posterity. 2 The census 
records which survive give detailed information of families with both 
parents' names. It is of course true that Egyptians, like all other peo- 
ple down to this day, could be tempted to falsify their genealogies or 

1 D. Henige, "Generation-counting and late New Kingdom chronology", JEA 67 
(1981), 182-184. 

2 M. L. Bierbrier, The Tomb-builders of the Pharaohs (London, 1982), 95— 6; J. Keith- 
Bennett, "Anthropoid Busts II", BES 3 (1981), 43-72. 


at any rate to attach themselves erroneously to a good family line so 
particularly illustrious ancestries should be treated with caution. The 
genealogical information becomes fuller as time progresses and is espe- 
cially voluminous from the late NK to the LP. Some pedigrees are 
known extending up to 13 generations, but it is usually difficult to fix 
these to a certain chronological range. 1 At least two pedigrees claim a 
genealogical link back to the MK. However, the inscription of Khnumibre c 
jumps abruptly from Dyn. 19 to Imhotep of Dyn. 3 and so cannot be 
taken as a serious link. 4 Similarly the inscription of c Ankhefensakhmet 
which names 60 generations, mostly holders of the high priesthood of 
Ptah, going back to the MK with the contemporary ruler for each, has 
too many gaps and inconsistencies to be accepted at face value.' It 
seems to have been put together from various lists of high priests (prob- 
ably unrelated) and other ancillary material. It can only be used when 
independent information is available from other sources. 

Another problem with genealogy is the lack of proper genealogical 
terms in the Egyptian language. "Father" jt, "Mother" mwt, "Son" zi, 
"Daughter" z?t, "Brother" sn, and "Sister" snt are used, but there are 
no words for other relationships so "uncle" is "brother of my father" 
or "nephew" is "son of my brother". 6 To avoid such long circumlo- 
cutions, jt / mwt could be used for "father- / mother-in-law", while sn 
might also mean "brother-in-law", "nephew" or even" uncle", or might 
conceivably be no relation at all but a good friend and contemporary. 
The term jt can also be used to specify an ancestor and not just a 
father.' One could disentangle the exact relationships if proper docu- 
mentation were available for all problems, but it is usually not. 

With regard to maternal relationships, it is not always clear that the 
wife of an official is the mother of his children unless specifically stated. 

3 J.-C. Goyon, "Les cultes d'Abydos a la basse epoque d'apres une stele du Musee 
de Lyon", Kemi 18 (1968), 29-44; R. El-Sayed, "Deux statues inedites du Musee du 
Caire", BIFAO 83 (1983), 135-143; L. M. Leahy & A. Leahy, "The Genealogy of a 
Priestly Family from Heliopolis", JEA 72 (1986), 133-147. 

4 Couyat & Montet, Ouddi Hammdmdt, Nos. 91-93; Wildung, Rolle, Doc. XVI. 130, 

' Borchardt, Mittel, 96-100; C. Maystre, Les grands pretres de Ptah de Memphis (Freiburg 
& Gottingen, 1992), 93-97. 

6 G. Robins, "The Relationships specified by Egyptian Kinship Terms of the Middle 
and New Kingdoms", CdE 54 (1979), 197-217; M. L. Bierbrier, "Terms of Relationship 
at Deir el-Medina", JEA 66 (1980), 100-107. 

7 G. A. Gaballa, The Memphite Tomb-chapel of Mose (Warminster, 1977), 22; Allam, 
Ostraka, 44. 


Such a relationship is usually assumed but may not be correct. The 
guardian Penbuy of Dyn. 19 is known to have had two wives — 
Amentetwosret and Iretnefer. A stela shows Penbuy, Iretnefer and his 
adult son Amenmose, but another shows him with Amentetwosret and 
an infant Amenmose. In the first stela, Amenmose must be shown with 
his step-mother, but the stela gives no indication of this. 8 The dissolu- 
tion of marriages by death or divorce would not have been uncom- 
mon, yet plurality of marriages is rarely documented. 

For genealogies to have a major impact on chronological problems, 
it is essential that there must be a clear and uncontroversial link of 
genealogy to some fixed chronological point, usually the reign of a 
King. If links can be found with successive generations, the effectiveness 
of the genealogy with regard to the order of succession and more impor- 
tantly the maximum passage of time becomes even more crucial. If the 
genealogy in question is that of the royal family itself, then the fixed 
links are self-evident. One important genealogy of this nature is that 
of Pasenhor of Dyn. 22 who traces his ancestry through 16 generations 
including four kings. 9 

Very little documentation, let alone genealogical information, sur- 
vives from the development from writing ca. 3400 BC to the end of 
Dyn. 3. One key document, recently discovered at Abydos, is the dynas- 
tic seal of King Den of Dyn. 1 which lists his immediate predecessors: 
Nar-mer, Aha, Djer, and Wadj. The genealogical content is provided 
at the end with the name of the King's mother as Merytneith. 10 In 
view of her appearance, it is highly probable that not only do we have 
a dynastic listing but also a genealogical listing — father-to-son. Unfor- 
tunately, the length of time per generation can only be estimated. A 
second seal of the last king of Dyn. 1 lists all his predecessors, but 
leaves out Merytneith. It would be optimistic to assume a straight father- 
to-son succession throughout the whole dynasty. No such documenta- 
tion survives for Dyns. 2—3 and here the order and genealogy of the 
kings remains uncertain. 

8 M. L. Bierbrier, Hieroglyphic Texts from Egyptian stelae etc. 10 (London, 1982), pi. 72; 
M. L. Bierbrier & H. de Meulenaere, "Hymne a Taoueret sur une stele de Deir el- 
Medineh," in R. Holthoer & T. Linders, eds., Sundries in honour of Torgny Save-Sb'derbergh 
(Uppsala, 1984), 23-29. 

9 Kitchen, TIP, 488, Table 19. 

10 G. Dreyer, "Ein Siegel der friihzeitlichen Kbnigsnekropole von Abydos", MDAIK 
43 (1987), 33-43; G. Dreyer, "Umm el-Qaab", MDAIK 52 (1996), 72-73. 


While the amount of information about private families increases 
from Dyn. 4, the tomb inscriptions do not normally name the parents 
of the deceased but more usually only his wife and children. Thus it 
is extremely difficult to put together a lengthy and coherent genealogy 
apart from those of the royal families of Dyns. 4—6. There too many 
gaps remain and many of the reconstructions are speculative. What is 
certain is that the genealogical succession to the throne when known 
did not run smoothly and did not follow in generation sequence in all 
cases so the king-lists alone do not reflect generations; for example, 
Khephren succeeded his brother Ra c djedef in Dyn. 4 and Ra'neferef 
was probably followed by his brother Neuserre c in Dyn. 5, while 
Nemtyemzaf I was followed by his brother Pepy II in Dyn. 6. Chronology 
here is indicated rather by the careers of these few officials who list 
the sovereigns that they served and so limit the time which has elapsed. 
Thus Sekhemkare c son of Khephren lived through the reigns from 
Khephren of Dyn. 4 to Sahure c of Dyn. 5 while Ptahshepses was 
brought up under Menkaure c and lived into the reign of Neuserre c . It 
has been suggested that Dyns. 3-4 and 4-5 were linked by marriages 
of royal heiresses, but there is no proof of this. 11 

Unfortunately, genealogy is not much of an aid in determining the 
chronology of the FIP. The genealogies of the royal families are uncer- 
tain and those of the nomarchs do not extend for enough time or pre- 
sent enough fixed links to prove crucial. The list of previous nomarchs 
in the tomb of Ukhhotep of Meir is unfortunately not complete or tied 
to any fixed points. Nor is it clear that the list is in any way genealogical. 12 

When adequate documentation occurs in the MK, there is still not 
enough material to present long coherent genealogies apart from the 
royal family of Dyn. 12 itself 13 The papyri from Kahun allow a few 
families to be reconstructed and the succession of office holders, in one 
case father-to-son, confirms the dating order of some papyri. 14 However 
these fragments are not crucial to the relative chronology of Dyn. 12 
which can be fixed by other methods. 

11 CAH 1/2, 145-189; M. Verner, Forgotten Pharaohs, lost pyramids: Abusir (Prague, 1994), 
134-155 on Ra'neferef. 

12 A. Blackmail, The Rock Tombs of Meir (London, 1915), III, 16-21. 

13 Franke, Personendaten. 

14 U. Luft, "Illahunstudien I: Zu der Chronologie und den Beamten in den Briefen 
aus Illahun", Oikumene 3 (1982), 101-156. 


Similarly no consecutive genealogies can be found during the SIP. 
Some detailed genealogies of short-lived royal families can be put 
together, but the links between them remain tenuous and open to 
different interpretations. It is possible that some kings indicated their 
filiation in their royal names which would provide a genealogical sequence 
but this view has perhaps been pushed too far. 15 The later genealogies 
which claim to reach back to the MK cannot be taken seriously and 
seem merely to look to well-known historical figures without any genealog- 
ical consistency. 

The advent of the NK led to an increasing amount of genealogical 
material in the form of statues, stelae, tomb inscriptions and papyri, 
both official and unofficial. The pedigrees of the royal families of Dyns. 
18-19 can be constructed in some detail apart from the confusion 
which occurs at the end of the dynasties. The regnal years known from 
other sources fix a chronology which does not conflict with the royal 
genealogies. The private genealogies and careers of officials again confirm 
the standard chronology that there are not too many years missing 
from the known regnal years. For example, c Ahmose Penekheb served 
from c Ahmose I to Hatshepsut so guaranteeing that the reigns of 
Amenhotep I and Thutmose I— II cannot extend over too long a period."' 

Until recently no family could be traced which extended from Dyn. 
18 to Dyn. 19 with the slight exception of the parents of Ramesses I 
who must have lived then but are only fixed in relation to their son 
and the dubious genealogy of the high priests of Ptah. New research 
has revealed that the high priest of Amun Wennefer lived at the end 
of Dyn. 18, certainly during the reign of Haremheb and possibly that 
of Tut'ankhamun. His younger son Amenmose is well attested in the 
reign of Ramesses II, being in charge of the Ramesseum probably from 
early in the reign when doubtless adult. The career of father and son 
prove that the contentious reigns of Haremheb and Sety I should not 
be unduly long but are still too imprecise to determine for certain a 
long or short reign for Haremheb." 

From Dyn. 19 onwards much more genealogical information becomes 
available concerning the royal court, the officials, and even humbler 

15 Ryholt, Situation, esp. 207-289. 

16 CAHll/l, 295. 

17 F. Kampp-Seyfried, "Die Verfemung des Namens P'-rn-nfi"; D. Raue, "Ein Wesir 
Ramses' II.", in: Stationen, 303-319 and 341-345. 


folk such as the workmen of Deir el-Medina. The royal inscriptions 
now name in detail the king's sons, daughters and wives, while the 
tombs of the officials and workmen give the names of the parents, 
grandparents, in-laws, and other relations. This new information can 
be crucial in determining the length of uncertain periods or the max- 
imum extent possible of reign lengths. For example, some uncertainty 
remains at the end of Dyn. 19 and the beginning of Dyn. 20 when it 
has been speculated in the past that an interregnum took place. The 
maximum year dates are known: Ramesses II, 66 years, 2 months (his 
last); Merneptah, 10 years, Sety II 6 years (his last), but it is not clear 
if the reigns of Amenmesses, 3 years; Siptah 7 years; Twosre 8 years; 
and Sethnakhte of Dyn. 20, 2 years are consecutive or contemporary 
with others. 18 However, certain individuals are attested who survive 
through this period into the reign of Ramesses III such as the vizier 
Hori, a great-grandson of Ramesses II 19 and the workman Nekhemmut 
son of Khons who is attested under Ramesses II and was a foreman 
in year 13 of Ramesses III. 20 The scribe Kenherkhepeshef is appar- 
ently attested at Deir el-Medina from year 30 of Ramesses II until the 
reign of Siptah, and his widow Naunakhte (who must have been a 
great deal younger than him) remarried, had eight children, and sur- 
vived until the reign of Ramesses V. 21 In view of this evidence, it seems 
logical that Merneptah could not have reigned much longer than ten 
years. It is highly probable that Amenmesse's years are to be totally 
contained within those of Sety II and that Twosre backdated her reign 
to the death of Sety II, thus encompassing that of Siptah and so reign- 
ing two years not eight. In view of Sethnakhte's recently discovered 
stela, he may well have dated his reign from the death of Siptah or 
shortly thereafter and so was contemporary and not consecutive with 
Twosre. More importantly, there can have been no interregnum and 
so the time elapsed from the death of Sety II to the accession of 
Ramesses III was probably not much more than ten years if that. 

18 If the highest known regnal date is the last, then the reign could have been as 
much as 1 1 months shorter. For the latest views on Siptah's reign see R. Drenkhahn, 
Die Elephantine- Stele des Sethnacht und ihr historischer Hintergrund (Wiesbaden, 1980); Jansen, 
Village Varia, 116. 

19 H. de Meulenaere, "Le vizier ramesside Hori", Annuaire de I'lnstitut de philologie et 
d'histoire orientates et slaves 20 (1968-72), 191-98. 

20 Bierbrier, Kingdom, esp. 30-33, corrected in M. L. Bierbrier, "The Family of Sen- 
nedjem", CdE 59 (1984), 199-213. 

21 Bierbrier, Kingdom, 26-29. 


A great deal of documentary evidence is available for the regnal 
dates of Dyn. 20, and genealogical information confirms that few extra 
years should be added to those known, certainly not more than a 
decade. The case of Naunakhte has already been mentioned, but other 
families from Deir el-Medina can be traced from Dyn. 19 to the begin- 
ning of Dyn. 21. The family of Sennedjem runs from Dyn. 19 to the 
beginning of Dyn. 21, notably Nekhemmut junior who is attested in 
office from Ramesses IV to Ramesses IX. The family of Kaha can be 
traced from early Dyn. 19 to Ramesses XL More importantly, the 
scribe Amennakhte was appointed to office in year 16 of Ramesses III 
and survived until Ramesses VI. His son is attested from Ramesses III 
to Ramesses IX, while his grandson appears from Ramesses VI to 
Ramesses IX. Three further generations are known through to Smendes 
of Dyn. 21. 22 The genealogies thus confirm that the extent of Dyn. 20 
is most likely correct, and there are no substantial gaps in chronology. 

From Dyn. 22 onwards a large body of texts on statuary and coffins 
record the genealogies of the royal families and the priestly class. Some 
of these extend back to Dyn. 21 and one even to Dyn. 19 although 
that genealogy is doubtful. 23 Although these genealogies are not pre- 
cise enough to solve the various chronological cruxes of the period, 
they do link Dyn. 21 to Dyn. 26, and even with the provision of a 
genealogical jump or two, they limit the time period so that it is unlikely 
that the entire time span can extend further then has been postulated 
by Kitchen. It may be slightly abbreviated, but again the genealogies 
make clear that the time span cannot be radically less than the gen- 
erally accepted chronology of the period. For example, the fourth 
prophet of Amun Djedkhonsuiuefankh, a contemporary of Osorkon I, 
has a great-great-grandson the fourth prophet of Amun Nakhtefmut, 
who was a contemporary of Osorkon III and whose granddaughter 
married into the Besenmut family which is attested from Dyn. 22 to 
Dyn. 26. 24 

22 Bierbrier, Kingdom, 19—44. 

23 Bierbrier, Kingdom, 51-53, where the line between Ipuy under Merneptah and 
Ankhefenkhons under Osorkon I is doubtful. 

24 For royal and priestly families of this period see Kitchen, TIP; Bierbrier, Kingdom, 
45-108, with revisions in Bibliotheca Orientalis 36 (1978), 306-309; G. Vittmann, Priester 
und Beamte im Theben der Spatzeit (Vienna, 1978); J. Taylor, "A priesdy family of the 
25th Dynasty," CdE 59 (1984), 27-57; D. Aston & J. Taylor, "The Family of Takeloth 
III and the 'Theban' Twenty-third Dynasty," in Leahy, Libya, 131—154. 


From Dyn. 26, the chronology of Egypt becomes firmly fixed to that 
of Persia, Greece, and ultimately Rome. Genealogy thus becomes for 
the most part unnecessary to determine chronological conundrums. 
However, family information can still prove useful in resolving some 
minor points at issue. Unfortunately, much of the genealogical infor- 
mation is patchy and uncertain with the most information from Thebes 
which was no longer the centre of major activity. Genealogies and indi- 
vidual careers can be constructed from surviving papyri and have been 
useful in confirming the dates of some of the obscure rebel pharaohs 
such as Harwennefer and 'Ankhwennefer. 25 The survival of coherent 
genealogical material from the Roman Period is severely limited and 
is no longer of any practical aid in settiing minor matters of chronology 
which remain. 

25 C. A. R. Andrews, Catalogue of Demotic Papyri in the British Museum IV. Ptolemaic Legal 
Texts from the Theban Area (London, 1990), passim for genealogies; P. W. Pestman, "A 
family archive which changes history", as well as, P. W. Pestman, "Haronnophris and 
Chaonnophris", in: Hundred-Gated Thebes: Acts of a Colloquium on Thebes and the Theban 
Area in the Graeco-Roman Period, S. P. Vleeming, ed., (Leiden: PLB 27, 1995), 91-137. 


The Editors 

During Dyn. 1 years were not yet counted, but simply named, as 
exemplified by the so-called year labels of Dyn. 1 that were attached 
to oil vessels. Since the reign of king "Serpent", labels were addition- 
ally inscribed with the hieroglyph for "year" (mpi). 1 Presumably epony- 
mous occasions were selected at the outset of or early in the year, being 
scheduled or symbolic and only coincidentally historical. 2 The labels 
are the earliest evidence for reckoning time by years, 3 but the type of 
year, that was in any case civil, not regnal, is not known. The annals 
attest that the "months" and days of [ c Aha's] last year and [Djer's] 
first year, add up to 10 "months" and 20 days, less than either a full 
lunar or solar year. The figures cannot be explained until further infor- 
mation comes to light; perhaps they are just a scribal error. 

Rnpt zm>-t>wy designated an incomplete accession year at least as early 
as [Djer] down through Dyn. 8. How the following years were expressed 
changed over time. At least from the penultimate year of [ c Aha] and 
through the end of Semer-khet's reign, i.e. for most of Dyn. 1, the 
annals record the biennial occurrence of "following of Horus", but there 
was no successive numbering of them. 4 Between Ny-netjer and Khase- 
khemwy at the end of Dyn. 2 the annals document a dating system 
that was based upon a regular biennial census {tnwt, jpi), 5 numbered 
successively within a reign and coinciding with the biennial "following 
of Horus". Whereas the "following of Horus" is still attested for Djoser, 
the biennial census is not. When it reappears under Snofru, it is not 
a regular biennial event. According to the annals, the 8th occurrence 

1 The earliest known "year label" dates from the reign of Narmer, see G. Dreyer, 
MDAIK 54 (1998), 138-139. 

2 Wilkinson, Annals, 63-64. 

3 Helck, Thinitenzeit, 144-175. 

4 Wilkinson, Annals, 90-91. 

5 Edel, Grammatik, 179-180. 


of the census followed the 7th occurrence without an intervening cen- 
sus-free year. Contemporaneous dates from Snofru's pyramid at Maidum 
attest a series of annual counts, recorded as rnpt zp x (year of the xth 
count), 6 interrupted now and then by a rnpt m-ht zp (year after the xth 
count).' Scholars are at odds about whether there was a regular bien- 
nial count thereafter during the OK; the annals of earlier Dyn. 6 indi- 
cate a biennial count. 1 ' 

Beginning in Dyn. 1 1 successive calendar years during each reign 
were counted. Gardiner presumed that during the MK the kings dated 
their second regnal year from the New Year's Day (I Akhet 1) follow- 
ing the actual day of their accession, so that their first year consisted 
merely of some months and days after the demise of their predecessor. 9 
This may have been the case, since pBerlin 10055 from Illahun attests 
that year 1 of [Amenemhet III] followed directly on year 19 of [Senwosret 
III], 10 which may reflect the accession of Amenemhat III after Senwosret 
III had died in the course of year 19 or 20. But it may also signify 
that a coregency began after year 19. 

In Dyn. 18 a new system was adopted which continued until at least 
ca. 800 BC. 11 Year 1 began on the actual day of accession and the 
following years were counted, accordingly, from the anniversary in the 
civil year, so that the civil year now always spanned parts of two reg- 
nal years. From Dyn. 26 through the Roman period the antedating 
system was introduced so that regnal year 2 began on I Akhet 1 after 
the accession. 12 

During the entire pharaonic period the concept of an "era" was 
employed only once. The whm mswt era began in 19 Ramesses XI and 
lasted at least 10 and perhaps as much as 12 years. 13 The so-called 
400 year era of Seth Nubti 14 should also be mentioned, as well as the 
attribution of the Amarna pharaohs' reigns to Haremhab. 1 ' During the 

6 For the reading rnpt zp, see E. Edel, JNES 8 (1949), 35-39; Gardiner read hit zp, 
see his Grammar, Excursus C. 
' See below, Chapter II. 4. 

8 See below, Chapter II. 5. 

9 Gardiner, "Years", 2 1—23, followed by Beckerath, Chronologic, 10. 

10 When Gardiner wrote, the date in pBerlin 10055 was mistakenly associated with 
Senwosret II and III. 

11 Cf. below Jansen-Winkeln, Chapter II. 10. 

12 A. Leahy, JEA 74 (1988), 187. 

13 See below, Chapter II. 8. 

11 Cf. Beckerath, Zjvischenzeit, 161. 
15 See below, Chapter II. 8. 


MK, the nomarchs counted their years of office independently from 
royal years, 16 and it is quite possible that the Theban HP's of early 
Dyn. 21 counted their pontifical years. 17 

The Solar Calendar of 365 Days 

There were 12 months of 30 days in the Egyptian calendar, supple- 
mented by five additional days, the so-called epagomenal or "added" 
days, for a total of 365 days. A tendency to regard the year as amount- 
ing to only 360 days is evident, for example when the daily income of 
a temple is stated to be one 360th of the yearly revenue. 18 The well- 
known disregard of the epagomenai in calendar schemes seems to be 
another consequence of this tendency. 19 

Dates within a year were expressed in terms of the three seasons of 
four months each: Akhet ('lit), "inundation"; Peret (pri) "winter", pre- 
sumably the season of the "emergence" (pr) of the fields from the flood; 
Shemu (stnw) "summer". 20 Originally months were numbered, not named, 
but a few month names are known from the MK. 21 Varying series of 
names are documented for the months in the NK from which the 
Greek, Aramaic, and Coptic names of the civil months derive; Ptolemy 
used the Greek names. 22 

The heliacal rising of the fixed star Sirius (Egyptian: Sothis) was the 
only astronomical event which occurred on a certain day in the 365- 
day civil calendar (see below, chapters III. 9.10). Because the calendar 
did not provide for an extra day every 4 years, the date of the rising 
of Sirius in the civil calendar shifted accordingly. The concept of a 
"Sothic cycle" for a complete shift of 1460 years (i.e., 365 days X 4) 
is first documented in the Hellenistic Period. The earliest examples for 
dating in the civil calendar of 365 days come from Djoser's reign, 23 or 
ca. 150 years after its introduction. 

16 Cf. for example W. K. Simpson, JARCE 38 (2001), 7-8, regnal year 43 of Senwosret 
I corresponding to year 25 of the nomarch Amenemhet. 
" See below, Chapter II. 9. 

18 Cf. G. A. Reisner, "The Tomb of Hepzefa, Nomarch of Siut", JEA 5 (1918), 84. 

19 Cf. for example, Leitz, Studien, 5-6; A. J. Spalinger, OLZ 87 (2002), 25. 

20 Gardiner, Grammar, Excursus C. 

21 J. P. Allen, The Heqanakht Papyri (New York: PMMAEE 27, 2002), 135-137. 

22 See below, Chapter III. 11. 

23 J. Kahl, N. Kloth & U. Zimmermann, Die Inschriften der 3. Dynastie (Wiesbaden: 
AA 56, 1995), 70-71. 


Calendar Adjustment 

R. Weill has played with the idea that the SIP was shorter than the 
standard chronology allows, implying that at a calendar adjustment 
days had been cut out of the 365-day year. 24 Parker argued against 
the feasibility of such an adjustment; 23 his arguments were reinforced 
by Depuydt. 26 

It is indeed correct that Pharaonic sources do not attest an interca- 
lated day. 27 But what if the Egyptians added or lost a calendar day in 
the course of their history, be it by intercalation or by mistaken day 
counting? The close agreement between late Egyptian and general 
chronology precludes such an error or deliberate intercalation after the 
beginning of the Saite Period. The loss or gain of a calendar day would 
shift the absolute date of a recorded lunar date by either exactiy -11 
(—14) years or by exactiy +14 (+11) years. 215 Shifts of this size are not 
compatible with the standard chronology of the NK. Furthermore the 
supposition of an intercalated day would break the link between 
the lunar date and a Sothic date implicit in the astronomical dates of 
the Illahun archive. 

It is well known that in 238 BC Ptolemy III attempted to introduce 
an improvement in the Egyptian calendar by adding a 366th day every 
four years, anticipating the later Julian calendar reform. But his edict 
went unheeded. A second attempt under Augustus succeeded in impos- 
ing a form of the Julian calendar on Egypt. This so-called Alexandrian 
calendar remained in use in Egypt until the Arab conquest, and it still 
survives today as the liturgical calendar of the Coptic and Ethiopian 

Alongside the 365 day calendar, a lunar calendar was also used. 
Specialists are not in accord on how it functioned in detail (see below, 
chapters III. 6.10). 

21 R. Weill, CdE 24/47 (1949), 13-18. 

25 R. A. Parker, RdE 9 (1952), 101-108. 

26 L. Depuydt, "On the Consistency of the Wandering Year as Backbone of Egyptian 
Chronology" JARCE 32 (1995), 43-58. 

27 Parker, Calendars, 39. 

28 It is left to the reader to figure that out. 


Conversion of Dates 

Astronomers reckon ancient astronomical events terms of the Julian, 
not the Gregorian calendar. The correlation between the Egyptian and 
the Julian calendars is implicit in Ptolemy's dating of astronomical obser- 
vations according to the Egyptian calendar and to the 365 day Egyptian 
year of the so-called Ptolemaic Royal Canon which began with year 
1 of the Babylonian king Nabonassar. 29 During the first years of the 
Era Nabonassar the correlation between the Egyptian and the corre- 
sponding Julian calendar dates was: 

Royal Canon, year 1 : Thoth 1 — February 26, 747 BC 

The Julian calendar year 745 BC was a leap year and so it follows 

Royal Canon, year 4 : Thoth 1 - February 25, 744 BC 
Royal Canon, year 8 : Thoth 1 - February 24, 740 BC 

On this basis it is possible to convert any Egyptian date in a given or 
chosen year into the Julian calendar equivalent, with due consideration 
that the Julian calendar has a leap year and the Egyptian calendar 
does not. Tables for conversion have been constructed by Neugebauer. 30 

Egyptian Calendar Day 

The word for calendar day used in dates was sw, not hrw (day as 
opposed to night). 31 Parker presented a circumstantial argument in favor 
of the beginning of the calendar day at dawn: "It is obvious that, when 
the [lunar] month begins, the first day of the month also begins." 32 
Most Egyptologists accepted the validity of Parker's argument. 33 However, 
in the 1980s Leitz and Luft revived the arguments presented by Sethe 
in the 1920s that the calendar day began with sunrise. 34 They were 

29 See below, Chapter III. 11. 

30 P. V. Neugebauer, Astronomische Nachrichten, No. 6261 (Kiel, 1937). 

31 Edel, Grammatik, 182. 

32 Parker, Calendars, 10. 

33 The validity of the argument was denied by Grzybek (n. 35), 142—143, who other- 
wise accepted the beginning of the calendar day at dawn. 

34 K. Sethe, "Die Zeitrechnung der alten Agypter III", NAWG (1920), 130-138. 
U. Luft, Altorientalische Forschungen 14 (1987), 3-11; Leitz, Studien, 1-5. 


countered by scholars like Grzybek, Spalinger and Wells who defended 
Parker's standpoint. 35 

Actually, August Bockh had solved the question in 1863 when he 
used Ptolemy's Almagest to demonstrate that the Egyptian calendar day 
began at dawn before sunrise. 36 In the Almagest Ptolemy twice recorded 
observations of Mercury, first with double dates: 

Hadrian, year 18, Epiphi 18 to 19, at dawn; 37 
Antoninus, year 4, Phamenoth 18 to 19, at dawn. 38 

Subsequentiy, he gave single dates for the same observations: 

Mercury as morning star: Hadrian, year 18, Epiphi 19 39 
Mercury as morning star: Antoninus, year 4, Phamenoth 19 40 

Mercury is only observable at dawn as morning star or at dusk as 
evening star. Ptolemy could have substituted single dates for double 
dates only if he reckoned dawn as the beginning of the Egyptian cal- 
endar days Epiphi 19 and Phamenoth 19. Accordingly, the Egyptian 
calendar day began during the observability of Mercury as morning 
star, i.e. at dawn before sunrise. Hieroglyphic sources also exist that 
imply the beginning of the calendar day at dawn — notably an entry in 
the Calendar of Lucky and Unlucky Days, describing the appearance 
of Seth in the bow of the solar bark which led Malinine to conclude 
that the calendar day began before sunrise, during hd-ti^ 1 The time 
when the god appeared at the bow of the solar bark, according to an 
entry in the Calendar of Lucky and Unlucky Days. 42 His argument is 
supported by the identification of the god Seth with planet Mercury 
here, rising shortiy before the sun and thus visible at dawn. 43 

35 E. Grzybek, Du calendrier macedonien au calendrier Ptolemaique (Basel: SBA 20, 1990), 
147-151; A.J. Spalinger, OLZ 87 (1982), 25; R. A. Wells, BiOr 49 (1992), 723. 

36 A. Bockh, Uber die vierjahrigen Sonnenkreise der Alten, vorziiglich iiber den Eudoxischen, 
(Berlin, 1863), 298-308.— Cf. R. Krauss, SAK 32 (2004), 275-286. 

37 J. G. Toomer, Ptolemy's Almagest, translated and annotated (Princeton, 1984), 449. 

38 Toomer (n. 37), 450. 

39 Toomer (n. 37), 456. 

40 Toomer (n. 37), 455-456. 

41 For hd-U as dawn, see K. Jansen-Winkeln, SAK23 (1996), 201-202, and D. Kurth, 
GM 108 (1989), 34-35. 

42 M. Malinine, "Nouveaux fragments du calendrier egyptien des jours fastes et 
nefastes", in: Melanges Maspero I, 2-3, Orient ancien (Cairo: MIFAO 66, 1935-1938), 
887-888, 898. 

43 R. Krauss (n. 36), 284-285. — For Seth as planet Mercury see Neugebauer & 
Parker, EAT III, 180. 


D-l dawn D dawn 

1 h— 

5 5+1 

7^ 7+ >- 

midnight midnight 

Fig. I. 5.1 

Because an Egyptian calendar day (D) begins at dawn before sunrise, 
it overlaps two Julian calendar days (8) and (8 + 1) that begin at mid- 
night as Fig. I. 5.1 shows. 

Neugebauer's tables yield 8 as equivalent of D. The assertion that 
an Egyptian calendar day D equals the Julian calendar day 8, means 
that the bright part of D corresponds to the bright part of 8. It is 
implied that the early hours of 8 are not contained in D, and that the 
last hours of D correspond to the first hours of the Julian calendar day 
(8 + 1). 



Stan Hendrickx 


There exists quite some confusion in the terminology used for the cul- 
tural phases and the relative chronology of Egypt during the 4th mil- 
lennium, generally known as the Predynastic period. Already the term 
"Predynastic" in itself may cause a problem. Although it evidently refers 
to all periods previous to the dynastic history of Egypt, its use is in 
reality reserved for the Naqada culture of (Upper) Egypt which dates 
to the 4th millennium BC and economically represents a late Neolithic 
culture with increasing social complexity which is at the origin of the 
pharaonic civilisation. Most authors also call the preceding Badari cul- 
ture "Predynastic", although it has also been considered "Neolithic", 
together with the 5th millennium finds from the Fayum, Merimde Beni 
Salama and el-Omari in LE. 2 

The terminology for the relative chronology of the Predynastic period 
is also far from consistent (Table II. 1.1). Originally Petrie distinguished 
three chronologically distinct cultures which he called respectively 
Amratian, Gerzean and Semainean, after type sites that he excavated. 3 
By doing so he stressed the material differences, especially those between 
the Amratian and Gerzean which according to Petrie were caused by 
the arrival of a "new race". This view has already been abandoned 
for a long time, but the terms Amratian and Gerzean in particular 
continue to be used by some scholars up to the present day. It is how- 
ever already a long time since the word "Naqada culture" has been 
used, divided in two or three phases (cf. below). This has the advan- 
tage of indicating the continuous development that took place during 
the 4th millennium BC. 

1 I wish to thank Werner Kaiser, Edwin van den Brink, Nathalie Buchez and 
Christiana Kohler for comments and information on various aspects of the problems 
concerned with this contribution. Joanne Rowland most kindly perused the English. 

2 E.g. B. Midant-Reynes, The Prehistory of Egypt. From the First Egyptians to the First 
Pharaohs (London 2000). 

3 Petrie, Diospolis Parva; idem, Prehistoric Egypt (London 1920). 



Table II. 1.1 

Petrie 1920: 46 


Scharff 1931: 




Hassan 1988 

Brunton 1937, 


Kantor 1944, 


1957, 1990 

1996, 1999 

Vandier 1953 


Baumgartel 1970 


Pro tody naslic 

Naqada Stufe 


Pre dynastic 


Naqada II 

Naqada Stufe 


Late Predynastic 


Naqada I 

Naqada Stufe 
la— c 


Middle Predynastic 





Early Predynastic 

In LE the local 4th millennium culture identified at Maadi and Buto, 
for example has recently been the focus of research. 4 The original name 
"Maadi culture/Maadian" is still used at present, 5 but it has meanwhile 
also been termed "Buto-Maadi culture" 6 and "Maadi-Buto culture". 7 
As up to now the site of Maadi has provided the largest amount of 
material available, it seems obvious that the terminology should stress 
the importance of that site. On the other hand, the position of "Buto" 
in the Maadi-Buto terminology indicates that the terminal stage(s) of 
this culture at Buto (and many other contemporary Delta sites) are no 
longer present at the key site of Maadi itself. On a more general level, 
it has been called "predynastic" because it is contemporaneous with 
part of the Naqada culture in UE, 8 but was also referred to as "chal- 

4 For recent overviews, see D. Fairings, "Ergebnisse der neuen Ausgrabungen in 
Buto. Chronologie unci Fernbeziehungen der Buto-Maadi-Kultur neu iiberdacht", in: 
Stationen, 35-45; idem, "Recent Excavations in Tell El-Fara'in/Buto: New Finds and 
their Chronological Implications", in: Eyre, ed., Proceedings, 365-375; idem, "The 
Chronological Frame and Social Structure of Buto in the Fourth Millennium BCE", 
in: Interrelations, 165-170. 

5 E.g. Midant-Reynes (n. 2). 

6 E.g. K. Schmidt, "Comments to the Lithic Industry of the Buto-Maadi Culture 
in Lower Egypt", in: L. Krzyzaniak, M. Kobusiewicz & J. A. Alexander, eds., Environmental 
Change and Human Culture in the Nile Basin and Northern Africa (Poznan 1993), 267-277; 

J. Wunderlich, T. von der Way & K. Schmidt, "Neue Fundstellen der Buto-Maadi- 
Kultur bei Esbet el-Qerdahi", MDAIK 4:5 (1989), 309-318; Fairings, in: Stationen, 35-45. 

' S. Hendrickx, "La chronologie de la prehistoire tardive et des debuts de l'histoire 
de l'Egypte", Archeo-Nil 9 (1999), 13-81; T. E. Levy & E. C. M. van den Brink, 
"Interaction Models, Egypt and the Levantine Periphery", in: Interrelations, 3-38. 

8 Midant-Reynes (n. 2). 


colithic" 9 or "late prehistoric". 10 More exceptionally it has been termed 
"Early Bronze", 11 mainly for comparison with the Levant. 

There is no obvious change in the material culture marking the tran- 
sition between Predynastic and Early Dynastic in UE. Here one should 
distinguish between cultural chronology based on material evidence and 
the historical chronology based on written documents. The latter how- 
ever are very exceptional outside of elite sites such as Abydos and 
Saqqara. Because of this, the end of the Naqada III culture is, from 
the archaeological point of view, to be placed within or at the end of 
Dyn. 2. The term Naqada culture will therefore also be used for the 
Early Dynastic period. 


A culture preceding the Badarian (cf. below) was originally identified 
by Brunton at Deir Tasa, and accordingly labelled by him as Tasian. 12 
One of the main archaeological characteristics are narrow beakers 
with flaring rim and incised decoration. Subsequently, the Tasian has 
been discredited by Baumgartel as a separate cultural entity and was 
considered part of the Badarian. 13 This view has been accepted for a 
long time, but more recently Kaiser has considered the Tasian as an 
entity distinct from the Badarian. 14 For him, the Tasian represents the 

9 F.-J. De Cree, '"Mutatis Mutandis' Egyptian Relations with Palestine in the 
Chalcolithic and Early Bronze Age I IV" GM 124 (1991), 21-42; K. Schmidt, "Tell 
el-Fara'in/Buto and el-Tell el-Iswid (south): The Lithic Industries from the Chalcolithic 
to the Early Old Kingdom", in: E. C. M. van den Brink, ed., The Nile Delta in Transition: 
4th.-3rd Millennium B.C. (Jerusalem, 1992), 31-42; idem, "Lower and Upper Egypt in 
the Chalcolithic Period. Evidence of the Lithic Industries: A View from Buto", in: L. 
Krzyzaniak, K. Kroeper & M. Kobusiewicz, eds., Interregional Contacts in the Later Prehistory 
of Northeastern Africa (Poznan, 1996), 279-289; S. P. Tutundzic, "Chalcolithic Canaan 
and Egypt: Reinvestigations and Considerations", Journal of the Serbian Archaeological Society 
12 (1996), 25-33. 

10 T. von der Way, Untersuchungen zur Spatvor- und Friihgeschichte Unteragyptens (Heidelberg: 
SAGA 8, 1993). 

11 A. R. Schulman, "At the Fringe: The Historiography and Historicity of the 
Relations of Egypt and Canaan in the Early Bronze Age I", in: P. R. de Miroschedji, 
ed., L'urbanisation de la Palestine a I'dge du bronze ancien (Oxford: BAR 527, 1989), 433-453. 

12 G. Brunton, Mostagedda and the Tasian Culture (London, 1937). 

13 E. J. Baumgartel, The Cultures of Prehistoric Egypt I (London, 1955), 20-21. 

14 W. Kaiser, "Zur Siidausdehnung der vorgeschichdichen Deltakulturen und zur 
fruhen Entwicklung Oberagyptens", MDAIK 41 (1985), 61-88. 


transmitter of LE elements to UE. However, this seems most unlikely 
because of the presence of beakers strongly resembling those of the 
Tasian in Sudanese neolithic contexts, for example at Kadero and 
Kadada, dated to late 5th— early 4th millennium cal. BC. 1 ' Most recently, 
Tasian beakers have also been published for the Eastern Desert. 16 In 
the Western Desert, Tasian burials have been identified in the Wadi 
el-Hol, on the desert road between Luxor and Farshut 17 and at Gebel 
Ramlah, near Nabta Playa. 18 The new evidence, together with a rein- 
terpretation of older information, allows rather for the consideration of 
the Tasian as a desert phenomenon, which however also had exten- 
sive contacts with the Nile valley. 19 The finds from Wadi Attula in the 
Eastern Desert have been radiocarbon dated between 4940 and 4455 
cal BC, a surprisingly early date. 20 These are the only dates available 
at present and although confirmation from other sites remains desir- 
able, this indicates not only a cultural difference with the Badarian but 
eventually also a chronological. It is however to be noted that the finds 
from the Nile valley occurring in a Badarian context, indicate that the 
Tasian may have started before the Badarian, but was at least also 
partially contemporaneous with it. 


The Badari culture has originally been identified in the region of Badari 
(Qaw el-Kebir, Hammamiya, 21 Mostagedda 22 and Matmar), 23 where a 
number of small cemeteries, containing in total about 600 tombs, and 

15 E.g. M. Chlodnicki, "New Types of the Neolithic Pottery in Kadero (Sudan)", 
CRIPEL 17,2 (1997), 29-35; J. Reinold, Kadruka and the Neolithic in the Northern 
Dongola Reach", Sudan & Nubia 5 (2001), 2-10. 

16 Friedman & Hobbs, "A 'Tasian' Tomb in Egypt's Eastern Desert", in: Gifts, 

17 J. C. Darnell & D. Darnell, "Opening the Narrow Doors of the Desert: Discoveries 
of the Theban Desert Road Survey", in: Gifts, 132-155. 

18 M. Kobusiewicz, J. Kabacinski, R. Schild, J. D. Irish & F. Wendorf, "Discovery 
of the first Neolithic cemetery in Egypt's Western Desert", Antiquity 78 (2004), 566-578; 
R. Schild et al., "Gebel Ramlah Playa", in: Tides of the Desert — Gezeiten der Wtiste, 
Jennerstrasse 8, eds., (Cologne, 2002), 117-23. 

19 Friedman & Hobbs (n. 16), 189. 

20 Friedman & Hobbs (n. 16), 178. 

21 G. Brunton & G. Caton Thompson, The Badarian Civilisation and Prehistoric Remains 
near Badari (London, 1928). 

22 Brunton (n. 12). 

23 G. Brunton, Matmar (London, 1948). 


40 poorly documented settlement sites. However, characteristic Badarian 
finds have also been made much further to the south (Mahgar Dendera, 24 
Armant, 25 Elkab, 26 and Hierakonpolis) 27 and also in the Wadi Ham- 
mamat. 28 Besides Mahgar Dendera most of these finds are unfortu- 
nately limited in extent. 29 

Until recently, most authors tended to consider the Badari culture 
as a chronologically separated unit, out of which the Naqada culture 
developed. However, the situation is certainly far more complex. Since 
more Badarian finds, or Badari related finds, have been made south 
of Badari, the Badari culture might well have been present between at 
least the Badari region and Hierakonpolis. Regional differences may 
have existed, the unit in the Badari region itself being the only well 
documented one up to now. 

The chronological position of the Badari culture is still subject for 
discussion. Its relative position as being older than the Naqada culture 
has already for a long time been shown by the stratified site at Hamma- 
miya. 30 From TL-dating, the culture might already have existed by 
5000 cal BC. However, based on the available radiocarbon dates, only 
the period around 4400-4000 cal BC can be confirmed for certain. 31 

Additional information on the chronological position of the Badarian 
may be gained by searching for its origin. Recent investigations have 
shown strong links between the ceramic industry of the Badarian and 
the Bashendi B unit from the Dakhleh oasis, 32 and with the Late/Final 
Neolithic from the Nabta-Kiseiba area. 33 The Bashendi B culture can 

24 S. Hendrickx, B. Midant-Reynes & W. Van Neer, Mahgar Dendera 2 (Haute Egypte), 
un site d'occupation Badarien (Leuven, 2001). 

25 O. H. Myers & H. W. Fairman, "Excavations at Armant 1929-1931", JEA 17 
(1931), 228-229. 

26 P. M. Vermeersch, Elkab II. LElkabien, Epipaleolithique de la Vallee du .Nil Egyptien 
(Leuven, 1978), 139-143. 

27 M. A. Hoffman, "A Preliminary Report on 1984 Excavations at Hierakonpolis", 
NARCE 132 (1986), 3-14. 

28 F. Debono, "Expediton archeologique royale au desert oriental (Keft — Kosseir): 
Rapport preliminaire sur la campagne 1949", ASAE 51 (1951), 74. 

29 For a more detailed overview of Badarian sites, see S. Hendrickx & E. C. M. 
van den Brink, "Inventory of Predynastic and Early Dynastic Cemetery and Settiement 
Sites in the Egyptian Nile Valley", in: Interrelations, 346—399. 

30 Brunton & Caton Thompson (n. 21), 73-8; D. L. Holmes & R. F. Friedman, 
"Survey and Test Excavations in the Badari Region, Egypt", PPS 60 (1994), 105-142. 

31 Hendrickx (n. 7), 19. 

32 C. A. Hope, "Early and Mid-Holocene Ceramics from the Dakhleh Oasis: Traditions 
and Influences", in: Gifts, 39—61. 

33 See especially site E-75-8, K. Nelson, "Ceramic Assemblages of the Nabta-Kiseiba 


be dated between 5650/5400 and 3950 cal BC, 34 and the Late/Final 
Neolithic at site E-75-8, roughly between 5500 and 4700 cal BC. From 
these dates a very early start for the Badarian seems distinctly possi- 
ble, contrarily to the opinion of the present author expressed only a 
few years ago. 3,1 Nevertheless it remains a fact that a duration of a 
thousand years or even more for the Badarian does not seem to be 
supported by the limited number of cemeteries and tombs known in 
the Badari region, where research is supposed to have been as exhaus- 
tive as possible. If originally based in its present desert regions, the 
Badarians might only at a later stage have used the Nile valley for per- 
manent living. 

JVaqada Culture — History of Research — Petrie's Sequence Dating 

In 1895 a huge cemetery of previously unknown type was discovered 
by W. M. F. Petrie at Naqada. 3b At first it was thought to date from the 
FIP, but Jacques de Morgan soon realised the prehistoric nature of the 
cemetery, 3 ' which was later confirmed as "predynastic" by Petrie. A 
number of important similar cemeteries were excavated in UE during 
the beginning of the last century. The original study on the relative 
chronology of the Naqada culture goes back to the early years of the 
20th century, when Petrie worked out his Sequence Dating, 38 the first 
attempt at what is now known as seriation. The history of this remark- 
able relative chronology has in recent years already been presented and 
discussed many times 39 and will only be presented briefly here. 

Area", in: K. Nelson & Associates, Holocene Settlement of the Egyptian Sahara. Volume 2. 
The Pottery ofJVabta Playa (New York— Boston— Dordrecht, 2002), 34-35. 

31 M. M. A. McDonald, "The Late Prehistoric radiocarbon Chronology for Dakhleh 
Oasis within the wider environmental and cultural Setting of the Egyptian Western 
Desert", in: C. A. Marlow & A.J. Mills, eds., The Oasis Papers 1: The Proceedings of the 
First Conference of the Dakhleh Oasis Project (Oxford, 2001), 26-42; idem, "Dakhleh Oasis 
in Predynastic and Early Dynastic Times: Bashendi B and the Sheikh Muftah Cultural 
Units", Archeo-Nil 12 (2002), 109-120. 

35 Hendrickx (n. 7), 19. 

36 W. M. F. Petrie & J. E. Quibell, Naqada and Ballas (London, 1896); E.J. Baumgartel, 
Petrie's Naqada Excavation: A Supplement (London, 1970). 

:i/ J. de Morgan, Recherches sur les origines de I'Egypte. I. L'age de la pierre et des metaux 
(Paris, 1896). 

:i8 W. M. F. Petrie, "Sequences in Prehistoric Remains", JRAI 29 (1899), 295-301; 
idem (n. 3), Diospolis Parva, 4-12; idem, Prehistoric Egypt (n. 3), 3-4. 

39 J. C. Payne, "The Chronology of Predynastic Egyptian Decorated Ware", Eretz- 



The Sequence Dating is based on the grave goods from the ceme- 
teries excavated by Petrie and his assistants at Naqada, Ballas 40 and 
Diospolis Parva. 41 As a first step, the pottery was arranged in a cor- 
pus of "predynastic" pottery, consisting of nine classes of pottery and 
over 700 types (Table II. 1.2). 42 Next, all of the objects from each grave 
were noted on a slip of card. Finally, the cards were arranged in a 
relative chronological order based on the resemblance of types. In this 
stage of his work, Petrie used only 900 relatively intact graves con- 
taining five or more different pottery types, out of over 4000 excavated 
graves. The chronological order was defined by two main principles. 
Firstly, an earlier and a later phase were distinguished through the 
observation that the classes of White Cross-lined pottery on one hand, 
and Decorated and Wavy Handled pottery on the other hand, hardly 
ever occurred together. Secondly, it was accepted that there had been 
an evolution of the shape of the Wavy Handled types, going from glob- 
ular to cylindrical shapes, while at the same time the handles had 
changed from two functional handles to a continuous decorative line. 41 

Table II. 1.2. Pottery Classes Distinguished by W. M. F. Petrie, Corpus of 
Prehistoric Pottery and Palettes (London, 1921) and Their Characteristics 


Defined by 

Class relations 




Firing technology 

Nile silt 



Surface treatment 

Nile silt 




mainly Red-Polished 

mainly Nile silt 

White Cross-Lined 




Nile silt 

Incised Black 



Nile silt 




mainly Late 

Marl clay 



Morphological detail 

Marl clay 



Fabric/surface treatment 

Nile silt, organic 




some Rough 

mainly Marl clay 

Israel 21 (1990), 77-82; D. C. Patch, The Origin and Early Development of Urbanism in 
Ancient Egypt: A Regional Study, U.M.I. (Ann Arbor / Pennsylvania, 1991), 153—170; 
B. Midant-Reynes, Prehistoire de lEgypte. Des premiers hommes aux premiers pharaons (Paris, 
1992), 240-243; idem (n. 2), 257-259; B. Adams, Ancient Nekhen. Garstang in the City of 
Hierakonpolis (New Maiden, 1995), 21-26; T. A. H. Wilkinson, State Formation in Egypt. 
Chronology and Society (BAR 651, Oxford, 1996), 9—11; S. Hendrickx, "The Relative 
Chronology of the Naqada Culture: Problems and Possibilities", in: A. J. Spencer, ed., 
Aspects of Early Egypt (London, 1996), 36-43; idem (n. 7), 20-25. 

40 Petrie & Quibell (n. 36). 

41 Petrie (n. 3), Diospolis Parva. 

42 W. M. F. Petrie, Corpus of Prehistoric Pottery and Palettes (London, 1921). 

43 Cf Petrie (n. 3), Diospolis Parva, pi. II; B. Adams, Predynastic Egypt (Aylesbury, 
1988), 27; Hendrickx (n. 7), 31, fig. 9. 


This evolution was corroborated by the fact that the "early" and "late" 
Wavy Handled types did not occur in the same tomb. In addition, the 
"late" Wavy Handled types and an important number of types from 
the Late class occurred in tombs which could be dated by inscriptions 
to Dyn. or 1. 

When all of the grave cards had been arranged in order, Petrie 
divided the cards into fifty equal groups, each of them consisting of 1 8 
graves, numbering them as Sequence Dates from thirty to eighty. By 
choosing to start at SD 30, he left space for earlier cultures, which he 
thought were still to be discovered. 44 Finally the fifty sequence dates 
were divided into three groups which he considered to be archaeolog- 
ically, culturally and chronologically different. The "cultures" were 
named Amratian (SD 30—37), Gerzean (SD 38—60) and Semainean (SD 
60-75), after important predynastic cemetery sites. 

The Sequence Dates were continued with a second typological cor- 
pus, for the "protodynastic" pottery. 43 This is almost exclusively based 
on material from the extensive cemeteries at Tarkhan. 46 This time the 
number of types reached 885 and no classes of pottery were distin- 
guished, which makes the corpus in some cases difficult to use. The 
"protodynastic" corpus partially overlaps with the most recent types of 
the "predynastic" corpus, as a result of which the Sequence Dates for 
the "protodynastic" corpus start already from SD 76 and continue to 
SD 86, which should mark the beginning of Dyn. 3. However, the SDs 
83-86 remained almost completely theoretical because of the lack of 
Dyn. 2 material at Tarkhan. The distinction between the individual 
Sequence Dates is not carried out in the way in which it was done for 
the "predynastic" corpus. This time, however, the transition to a new 
SD is based on typological breaks which Petrie defined mainly through 
the development of the Wavy Handled types. Finally, Petrie connected 
the Sequence Dating with the historically dated pottery types and other 
objects from the royal tombs of the earliest dynasties at Abydos. 4 ' 

44 When he discovered the Badarian, Brunton tried to apply Sequence Dating 
(Brunton & Caton Thompson (n. 21), 26; Brunton (n. 12), 50-51), but this was from 
the beginning considered problematic and never found its way into the scientific 

45 W. M. F. Petrie, Corpus of Proto-Dynastic Pottery (London, 1953). 

46 W. M. F. Petrie, Tarkhan I and Memphis V (London, 1913); idem, Tarkhan II 
(London, 1914). 

4/ Petrie (n. 46), Tarkhan I and Memphis V, 3. 


The development of the Sequence Dates certainly represents one of 
the major intellectual performances in the study of predynastic Egypt, 
and most of the basic observations by Petrie, such as the evolution of 
the Wavy Handled types, were never contradicted. Nevertheless, a num- 
ber of methodological shortcuts and possible errors concerning the prac- 
tical elaboration of the Sequence Dates were subsequently pointed out 
by several authors. 48 Petrie makes no clear distinction between typol- 
ogy and chronology. This, for example, is obvious from the heteroge- 
neous manner by which his pottery classes have been defined (cf. Table 
II. 1.2). Furthermore, the definition of the individual types within these 
classes is not bound by strict rules. 49 Even more important is the fact 
that Petrie only included tombs with five or more objects, resulting in 
the under-representation of the earlier period. 50 The lack of attention 
to eventual regional differences is also to be noted. 

The most striking omission in Petrie 's manner of working remains 
that he never took the horizontal distribution of the graves into con- 
sideration. This, despite the fact that he noted for instance that none 
of the cemeteries from Diospolis Parva covered the whole of the SDs, 
but that "early" and "late" cemeteries could be distinguished. 51 Strangely 
enough, Petrie does not mention spatial distribution within the ceme- 
teries of Naqada, Ballas or Diospolis Parva, although it is hardly imag- 
inable that he did not notice any clustering of tombs with similar 
funerary equipment. On the occasion of later excavations, by former 
assistants of Petrie, the existence of groups of chronologically related 
graves, and therefore the differences in the spatial distribution of objects, 
were noticed several times and at different sites, 52 but no attempts were 
made to use these observations for chronological purposes. 

48 G. F. Legge, "New Light on Sequence-dating", PSBA 35 (1913), 101-113; A. Scharff, 
Das vorgeschkhtliche Grdberfeld von Abusir el-Mekq (Leipzig, 1926), 71-74; H. J. Kantor, 
"The Final Phase of Predynastic Culture, Gerzean of Semainean?", JNES 3 (1944), 
110-136; Baumgartel (n. 13), 2; idem, Predynastic Egypt (CAH I, IXa, London, 1970), 
3-5; W. Kaiser, "Stand und Probleme der agyptische Vorgeschichtsforschung", £AS 81 
(1956), 87-109; Hendrickx (n. 39), 37-38; idem (n. 7), 21-22. 

49 Petrie (n. 42), 5. 

50 Hendrickx (n. 39), 37. 

51 Petrie (n. 3), Diospolis Parva, 31-32. 

52 D. Randall-Mclver & A. C. Mace, El Amrah and Abydos. 1899-1901 (London, 
1902), 3; E. R. Ayrton & W. L. S. Loat, Pre-dynastic Cemetery at El-Mahasna (London, 
1911), 2; T. E. Peet, The Cemeteries of Abydos. Part II. 1911-1912 (London, 1914), 18; 
Brunton & Caton Thompson (n. 21), 50—1. 


Finally, Petrie's wish for a very detailed relative chronology in 50 
Sequence Dates causes a fundamental problem. As a basic principle, 
the definition of the original Sequence Dates was made in a manner 
so as to minimise the chronological dispersion of each type of pottery. 
This results in a compromise between the competing claims of all pot- 
tery types for closer proximity. However, the "perfect balance" obtained 
by Petrie is purely artificial, since, whenever new graves will be added 
to the system, the range of Sequence Dates for a number of types will 
have to be expanded, and the accuracy suggested by the Sequence 
Dating system becomes purely hypothetical. The integration of new 
cemeteries over time made the whole system more and more problematic. 

Kaiser's Stufen Chronology 

W. Kaiser was the first to reinvestigate the relative chronology of the 
predynastic period in a fundamental manner.' 13 He used the horizontal 
distribution of pottery classes and types of objects within cemetery 
1400—1500 at Armant as point of departure. 54 Three spatial zones were 
distinguished by the relative percentages of Black-Topped, Rough and 
Late Wares, each of them dominating one zone. These zones are con- 
sidered to represent chronological stages. Within these three periods, 
subdivisions, called Stufen, were recognised according to the clustering 
of types of objects, these being almost exclusively pottery. The results 
of the analysis of the Armant cemetery are completed with a limited 
investigation of cemeteries for which the publication was less detailed, 
but where pottery types occur which are not represented at Armant. 
In this manner Kaiser distinguished three main stages of the develop- 
ment of the Naqada culture, each with their subdivisions. All in all, 
1 1 Stufen are identified, the two earliest and the two most recent of 
which are not represented at Armant. The archaeological description 
of the Stufen is based on types of objects, according to the Petrie typol- 
ogy, which Kaiser accepts as characteristic for a particular Stufe. Material 
from cemeteries other than Armant is also included. The characteristics 
of the Stufen are also used by Kaiser to study the geographical distri- 
bution of the Naqada culture. The chronological expansion which can 

53 W. Kaiser, "Zur inneren Chronologie der Naqadakultur", Archaeologia Geographica 
6 (1957), 69-77. 

54 R. L. Mond & O. H. Myers, Cemeteries of Armant I (London, 1937). 



Table II. 1.3. Correlation between Petrie's Sequence Dating and Kaiser's Stufen 





Zeitstufe III 

SD 63-80 


SD 63-76 

Zeitstufe lied 

SD 40/45-63 


SD 38-62 

Zeitstufe Ilab 

SD 38-40/45 

Zeitstufe I 

SD 30-38 


SD 31-37 

be observed from UE both towards the north and the south, is con- 
sidered further proof of the validity of the Stufen chronology. 

The Stufen have also been compared with the Sequence Dating.' 5 
Although Kaiser defined three main periods, the SD's attributed to 
them show that they are largely, but not completely, identical to the 
Amratian, Gerzean and Semainean distinguished by Petrie (Table II. 
1.3). When compared to the Sequence Dating, Kaiser's system has the 
advantage of including not only information from the typological appa- 
ratus, but also from the spatial distribution of the objects. Furthermore 
it does not give the idea of extreme accuracy, but by defining periods, 
it escapes largely, although not completely, the problem of becoming 
increasingly meaningless when new data are added. 

However, this does not mean that the Stufen system is without its 
problems. Although Kaiser included data from a number of cemete- 
ries besides the one at Armant, it remains a basic fact that data from 
only a single cemetery are used for the description of the Naqada cul- 
ture throughout UE. Nevertheless, Kaiser is well aware of the possi- 
bilities for regional differentiation, and has noticed regional phenomenon 
at Mahasna for instance. 56 The problem caused by using the cemetery 
at Armant becomes even more complicated because the earliest phase 
of the Naqada culture is not present at Armant, and even the most 
recent phases are very sparsely documented or absent. Therefore, the 
definition of the Stufen la and lb is merely based on hypothesis, although 
examples from other cemeteries besides Armant are given. The descrip- 
tion of Stufe Illb, though less hypothetical than Stufen la and lb, is also 
based on information from other cemeteries. In most cases it was not 
possible to study the spatial development of these cemeteries, and there- 
fore Kaiser's description of the Stufen la— b and Illb depends largely on 

55 Kaiser (n. 48), 109. 

56 Kaiser (n. 53), 74. 


the theoretical evolution of pottery types, especially from the Wavy- 
Handled class as already accepted by Petrie. 

A very practical problem is that Kaiser's study was only published 
in abridged version as an article. Because of spatial considerations within 
the publication, Kaiser was unable to provide details on his analytical 
method. The plates illustrating the article visualise the description of 
the Stufen by presenting for each of them the most important and 
characteristic types of objects.' 7 Unfortunately these plates have been 
used by some as absolute guidelines, despite the fact that the plates are 
only intended to be considered as an idealised outline of the develop- 
ment of the Stufen. This created a false idea of secure dating, especially 
for the Stufen Ilia— b. More recently Kaiser mentioned in an article the 
extension of his Stufen chronology into Dyn.l. 58 However, the manner 
in which this was done remains until now unpublished. The extension 
was nevertheless used by several authors. 

Kaiser's distinction of three phases within the Naqada culture reflects 
Petrie's original division. At first view there seems to be no problem 
because several cemeteries belonging to the Naqada culture bear evi- 
dence for the presence of three groups of graves, dominated respec- 
tively by the presence of Black-Topped, Rough and Late/Wavy Handled 
pottery. The moment at which the transitions are placed is however 
not beyond dispute. This and other more specific problems concern- 
ing the Stufen chronology will be discussed when the individual subdi- 
visions of the Naqada culture are presented. 

Computer Seriation 

Computer seriation has also been applied on predynastic cemeteries to 
study their relative chronology. A pioneer attempt made by E. M. 
Wilkinson is at present only of historical interest. 59 Far more important 
is Kemp's seriation by multi-dimensional scaling of the graves within 
cemetery B at el-Amrah and the cemetery of el-Mahasna. 6() However, 

57 Kaiser (n. 53), PI. 21-24. 

58 W. Kaiser, "Zur Entstehung des gesamtagyptischen Staates", MDAIK 46 (1990), 
Abb. 1. 

59 E. M. Wilkinson, "Techniques of Data Analysis. Seriation Theory", Technische und 
Naturwissenschqftliche Beitrage zur Feldarchaologie. Archaeo-Physika 5 (1974), 1-134. 

60 B. J. Kemp, "Automatic Analysis of Predynastic Cemeteries: A New Method for 
an Old Problem", JEA 68 (1982), 5-15. 


this seriation is not used for the evaluation of Kaiser's Stiffen chronol- 
ogy, but for Petrie's Sequence Dating. 

A far more elaborate study of the relative chronology using seriation 
has been made by T. A. H. Wilkinson. 61 Eight Predynastic — Early 
Dynastic cemeteries were seriated. 62 For this purpose, 1420 out of 1542 
types from Petrie's corpus which occurred in the eight cemeteries were 
condensed into 141 groups. 63 This approach, of course, carries the risk 
of producing strongly heterogeneous types, and as Wilkinson himself 
notes, some of his groups "bring together types with broad similarities 
but some significant differences". 64 There are indeed problems with the 
majority of the newly defined groups, especially for the plates, cups 
and bowls (P 001, P 004, P 033, P 034 etc.). Among the most notable 
difficulties is the grouping of Nile silt and marl clay bowls and even 
the very characteristic Nile silt bread moulds in the groups P 094, 
P 095 and P 103. Several groups of jars are also very heterogeneous 
(e.g. P 013, P 019, P 029, P 128, PI 31, PI 38). One can also question 
the validity of distinguishing two groups of wine jars (P 107 and P 108) 
by the broadness of their shoulders, while very distinctive elements such 
as the applied ridges and the wavy decorations are completely ignored. 
Furthermore, it is difficult to understand why in group P 052 cylin- 
drical jars both with and without incised decoration have been grouped. 
This is all the more disturbing because this element has been used as 
a chronological indicator by Kaiser for the development of his Stufen 
chronology. It is obvious that the manner in which Petrie's types have 
been grouped, makes it impossible to arrive at the same results as those 
which Kaiser obtained for his Stufen chronology. 

The inevitable conclusion is that when Petrie's types are grouped 
into only 141 new groups, these become heterogeneous to a degree no 
longer consistent with the concept "type". While this does not render 
Wilkinson's seriations totally meaningless, the many anomalies involved 
introduce a disturbing element of uncertainty in the results. 6 ' 1 Whenever 

61 T. A. H. Wilkinson, "A New Comparative Chronology for the Predynastic — Early 
Dynastic Transition", JACF 7 (1994-1995), 5-26; idem (n. 39). 

62 Tarkhan Hill and Valley Cemeteries, Turah, Matmar Cemeteries 200/3000- 
3100/5100, Mostagedda Cemetery 1600-1800, Mahasna, el-Amrah Cemetery b, Armant 
Cemetery 1400-1500, Hierakonpolis Fort Cemetery. 

63 Wilkinson (n. 39), 25. 

64 Wilkinson (n. 39), 23. 

65 For a more extensive discussion of Wilkinson, State Formation in Egypt (n. 39), see 
the review by S. Hendrickx in JEA 85 (1999), 241-245. 


possible, the result of the seriations has also been plotted on a plan of 
the cemetery, to facilitate discussion of the chronological development 
of the cemeteries involved. The horizontal distribution of graves and 
pottery types is, however, only used as a method of control and not 
as a primary source of information. The use of a general typology for 
all eight cemeteries allowed Wilkinson direct comparison of the seri- 
ation results. Each sequence distinguished was compared against Kaiser's 
Stufen chronology. Significant differences emerged, most notably in 
Kaiser's demarcation of the three major Naqada culture phases. 

Distribution Studies 

A number of studies on the relative chronology of the Naqada period 
have started from the spatial distribution of objects within the ceme- 
teries rather than from seriation. Once more, the problem of group- 
ing types, which was the main problem for seriation, will occur, albeit 
to a lesser extent. The first study of this kind concerns an unpublished 
M.A. thesis by R. Friedman on the spatial distribution and relative 
chronology at Naqa ed-Deir cemetery 7000. h6 Comparison is made with 
Kaiser's Stufen chronology. Spatially distinguished groups of graves with 
objects characteristic for the Stufen Ic-IId are also represented at Naqa 
ed-Deir, but some differences in the characteristic pottery types for the 
individual Stufen have also been observed. 

J. C. Payne applied Kaiser's chronology to the available information 
for the Main Cemetery at Naqada. 6 ' She concludes that the same Stufen 
can be distinguished both at Armant and at Naqada and also that the 
differences in the archaeological description of the Stufen remain very 
limited, the most important being situated in Stufe IIb. 6!i 

M ' R. F. Friedman, Spatial Distribution in a Predynastic Cemetery: Naqa ed Der 7000 (unpubl. 
MA. diss., Berkeley, 1981). The chronological framework established by Friedman was 
afterwards used by S. H. Savage, Descent, Power and Competition in Predynastic Egypt: Mortuary 
Evidence from Cemetery N 7000 at Naga-ed-Der (U.M.I., Ann Arbor/ Arizona State University, 
1995); idem, "Descent Group Competition and Economic Strategies in Predynastic 
Egypt", JAnAr 16 (1997), 226-68. For the latter study see however P. Delrue, "The 
Predynastic Cemetery N7000 at Naga ed-Der. A Re-evaluation", in: H. Willems, ed., 
Social Aspects of Funerary Culture in the Egyptian Old and Middle Kingdoms (Leuven, 2001), 

( " J. C. Payne (n. 39), 77—82; idem, "Predynastic Chronology at Naqada", in: R. F. 
Friedman & B. Adams, eds., The Followers of Horus. Studies Dedicated to Michael Allen 
Hoffman (Oxford, 1992), 185-192. 

68 Payne (n. 39), 81. 


The most extensive study based on spatial distribution is a still incom- 
pletely published Ph.D. dissertation by the present author. 69 The limited 
number of "Predynastic" cemeteries for which both a map and a grave 
register, be it eventually incomplete, are available served as a starting 
point.' For the Early Dynastic period information came from a number 
of "Protodynastic" and "Archaic" cemeteries in LE.' 1 Methodologically, 
there is not much difference to the method already developed by Kaiser. 
This implies that the distinction of related groups of graves is not only 
based on their contents but at the same time on their spatial distribu- 
tion within the cemetery. As a result, a conflict of interests will arise 
between the search for closer chronological proximity of all examples 
of one pottery type on one hand, and the definition of spatially well 
defined groups of graves on the other hand. Neither one of these two 

69 S. Hendrickx, De grajvelden der Naqada-cultuur in ^uid-Egypte, met bijzondere aandacht 
voor het Naqada III grqfveld te Elkab. Interne chronologie en sociale differentiate (unpubl. PhD. 
diss., Leuven, 1989); idem (n. 39); idem (n. 7). 

70 The following cemeteries, from north to south, are involved: Matmar, cemetery 
2600-2700 [Brunton (n. 23), pi. VIII-IX, XIX], el Badari, cemeteries 3700 and 3800 
[Brunton & Caton-Thompson (n. 21), pi. Ill, XXXII-XXXIII] , Hammamiya, ceme- 
tery 1500-1800 [G. Brunton, Qau and Badari I (London, 1927), pi. VI, X XI; Brunton 
& Caton-Thompson (n. 21), pi. XXX-XXXI]; Qaw el-Kebir, cemetery 100 [idem, pi. 
Ill, XXX], el-Salmany [A. el Sayed, "A Prehistoric Cemetery in the Abydos Area", 
MDAIK 35 (1979), 249-301], Naqada Main Cemetery [Petrie & Quibell (n. 36), pi. 
LXXXII-LXXXIII; Baumgartel Petrie's Naqada Excavation: A Supplement, S. Hendrickx, 
"Predynastische objecten uit Naqada en Diospolis Parva (Boven-Egypte)", Bulletin van 
de Koninklijke Musea voor Kunst en Geschiedenis 57 (1986), 31-44; J. C. Payne, "Appendix to 
Naqada Excavations Supplement", JEA 73 (1987), 181—190], Armant cemeteries 1300 
and 1400-1500 [Mond & Myers, n. 54], Hierakonpolis, locality 27, 'Fort Cemetery' 
[B. Adams, The Fort Cemetery at Hierakonpolis (London — New York, 1987)], Elkab 
[S. Hendrickx, Elkab V. The Naqada III Cemetery (Brussels, 1994)], Kubbaniya, South 
cemetery [H. Junker, Bericht iiber die Grabungen von der Akadamie der Wissenschqften in Wien, 
auf den Friedhofen von El Kubanieh — Sud. 1910-1911 (Wien, 1919)]. 

71 The following cemeteries, from north to south, are involved: Abu Roash, ceme- 
teries 0, 300, 400-500, 800-900 & M [A. Klasens, "The Excavations of the Leiden 
Museum of Antiquities at Abu-Roash. Report of the First Season: 1957. Part I", OMRO 
38 (1957), 58-68; idem, "The Excavations of the Leiden Museum of Antiquities at 
Abu-Roash: Report of the First Season 1957. Part II", OMRO 39 (1958), 20-31; idem, 
"The Excavations of the Leiden Museum of Antiquities at Abu-Roash: Report of the 
Second Season 1958. Part I", OMRO 39 (1958), 32-55; idem, "The Excavations of 
the Leiden Museum of Antiquities at Abu-Roash: Report of the Second Season 1958. 
Part II. Cemetery 400", OMRO 40 (1959), 41-61, idem, "The Excavations of the 
Leiden Museum of Antiquites at Abu-Roash: Report of the Third Season 1959. Part 
I", OMRO 41 (1960), 69-94; idem, "The Excavations of the Leiden Museum of 
Antiquities at Abu-Roach: Report of the Third Season 1959. Part II. Cemetery M", 
OMRO 42 (1961), 108-28], Saqqara, cemetery west of Serapeum [R. Macramallah, 
Un cimetiere archaique de la classe mcyenne du peuple a Saqqarah (Le Caire, 1940)], Turah 
[Junker, Turah], Tarkhan, Valley cemetery, cemeteries A, F, G, H, J & Q_ [Petrie 
(n. 46), Tarkhan I; idem (n. 46), Tarkhan II]. 


elements can be accepted as prevailing over the other. Thus, most 
unfortunately, it seems impossible to establish clearly defined, "objec- 
tive" rules for the definition of archaeological complexes representing 
relative chronological periods within the Naqada culture. This implies 
that although a relative chronology defined in this manner is of course 
based on the seriation principle, it nevertheless depends considerably 
on the personal interpretation of the researcher. 

Most cemeteries of the Naqada culture were used over several cen- 
turies, during which they expanded horizontally, but not in a pre- 
meditated manner. For the Naqada I— II period in particular this results 
in cemeteries consisting of "patches" of simultaneous tombs, as can be 
shown for the cemetery at Adai'ma for example.' 2 During the Naqada 
III period, the cemeteries tend to develop in a more linear manner, 
as can be seen at Elkab 73 and Tura. 74 By comparing the cemeteries 
that were analysed, it becomes clear that similar groups exist for different 
cemeteries. In this manner, 1 1 groups of graves, an equal number to 
Kaiser's Stufen, axe. distinguished and their relative chronological order 
defined through their mutual position in the cemeteries, and through 
the evolution of the pottery classes and types of objects. However, com- 
paring groups of related objects from geographically different cemeter- 
ies does not have to imply that they are also contemporaneous in 
absolute chronological terms. Unfortunately, this question cannot be 
answered because of the limited number of C 1 4 dates available for the 
Naqada cemeteries from UE. For this reason, and since related groups 
of archaeological objects can be distinguished at several sites, until 
proven otherwise, we may as well accept the simultaneity of archaeo- 
logical groups with a strong resemblance, suggesting that the same 
chronological periods may well have existed for the different cemeteries. 

In a further stage, the data from cemeteries without published maps 
were integrated and the possibilities of regional variability investigated. 
This eventually allows for an archaeological description of each of the 
relative chronological periods. The general observations made by Kaiser 
for cemetery 1400-1500 at Armant are not fundamentally contradicted, 

72 E. Crubezy, T. Janin & B. Midant-Reynes, Adaima II. La necropole predynastique (Le 
Caire, 2002), 415-417. 

73 Hendrickx (n. 70, Elkab V), 205-216. 

74 Hendrickx (n. 39), 57-59. 


and therefore the number of relative chronological periods is equal to 
the number of Stufen distinguished by Kaiser, although in some cases 
important differences occur in their archaeological description (cf below). 
It was therefore decided to follow Kaiser's work as closely as possible, 
but to replace the word "Stufe" by "Naqada" and at the same time 
change the letter indication into capital letters, which results in "Naqada 
IA" etc. 75 

Description of Naqada Periods — Naqada I 

The descriptions presented here are mainly based on the distribution 
studies made for a number of cemeteries for which both a map and 
tomb inventory are available. 76 Only the main developments and most 
characteristic types of objects are mentioned (cf. Table II. 1. 4a— b). 
Although the archaeological characteristics of the Badarian are strongly 
related to those of the early Naqada I period, it is nevertheless to be 
noted that at present no cemeteries are known which show a contin- 
uous use from the Badarian into the Naqada period. The burials are 
however of the same type and organised in the same manner. The 
differences are mainly to be found in the material culture, for which 
the pottery presents the most obvious evidence. Not only will the char- 
acteristic rippled surface of the Badarian only occur most exception- 
ally during Naqada IA, but more importantly, the three principal pottery 
fabrics of the Badarian differ from those of the Naqada culture.' 7 Naqada 
I sites occur only in UE, from the Badarian region in the north to a 
few sites south of Aswan, in Lower Nubia. All of the Naqada I period 
is dominated by Black-Topped pottery, which makes up over 50% of 
the assemblage. Red-Polished and White Cross-Lined are the second 
most important categories. 

75 Cf. Levy & van den Brink, in: Interrelations, 9. 

76 Hendrickx (n. 69); see notes 70—7 1 . For a more detailed overview of the relative 
chronological periods represented at individual Naqada sites, see Hendrickx & van den 
Brink, in: Interrelations, 346—399. 

" R. F. Friedman, Predynastic Settlement Ceramics of Upper Egypt: A Comparative Study of 
the Ceramics of Hemamieh, Nagada and Hierakonpolis (U.M.I., Ann Arbor / Berkeley, 1994). 







■«t ° 






3 o i£ 

o -h 


J3 I O 

< O O 

_i O CM 



C9 „ 

-'« CO 


■ o 

I * 

O c9, 

<° a 




°" G 

o g 

-H (J 

CO o 









o o 
•g cm 

u © 


T3 . 







r S 

C9 ^ 



S w 

-2 S 




.2 £c 





o ai 

o o 
\d co 

e§ o 

a S 


O O 

w O 


a b 

— u 




co r^. 

-a ° ° 

S CO co 

Ph .cm 

o „ 



CM c9 


O -~ 

^H O 

o <=> 

























i— i 












■fi ■ 

+- 1 ^ 




8 £ 


OJ ° 

H 2 



_d S 

'"j CJ 


2 a 

H -a 



-*— ' 


wT -<— ■ 

flJ b 

•a S 


cj _=q 

t; ""3 



<u -^ 


u § 

►. s 

<v ^ 

^3 rt 


-0 <u 

a ■" 


CG (9 



d oj 



3 a 

G g 


H ^a 


cd <u 

o -a 
co « 



o ° 

O cu 

r^ "3 


r (9 

O cj 

o <u 

(O jd 

o -^i>-cNrcT)CO^- | CT)Ln 

^h CM CO CM 

o r-- I m | ^h 
co r- -^ 


LT! N ffl CM lO 

CO CO O ^h — i 

— I -H -H ^ ^J0 -H 

CM -h 


O IMCO O) ^ ^ rH 


rH If) CO lO M rH IM 

— i CO 


CM ^-i t^« 

^ ^ cri m cm cm I co r^ o 
CM !wO ^^ in ^^ CM ^h 

CO (n I CMOO I ^ 

tD Tf< I -h CM 

r-i r-i CO 

CM ^ 

1 c^ 












1 o 









CM in CM 















— I CC -h 


CM ^ ^ 

r^ —i ,_- 



CM rf -h 














co" o" o" — " m" — r co" r-." cr>" r^ ^" o" 

tQ_ CD 

cm cr>" 







































^ CM lT^ CM iO^ <£^ CO^ — ^ 

o" cT o" <^o crT co" ^ o" 


o^ ^ co_ ^ co^ co^ r^ o^ cm 
o" o" cT co" co" i— r -tjT c^T cT 

-h CO ^ 








cr ; 






, , 













' ' 


' ' 


1 ' 









































to^ iO^ 

m" — " 
















co" r-." 












oo" r-~" 












to" to" 
to — 






muZ;hfcQrth-i^ s S-cS 


JVaqada IA 

The distinction made by Kaiser between Stufe la and lb cannot be 
confirmed beyond doubt for the cemeteries for which a map is available. /!i 
This was also noted by Friedman for cemetery 7000 at Naqa ed-Deir, 
where tombs predating Stufe Ic are present, but could not be distin- 
guished in groups matching Kaiser's Stufe la and lb. 79 Only in the east- 
ern part of the Naqada Main Cemetery was it possible to distinguish 
spatially a number of Naqada IA— B tombs, but a clear distinction 
between IA and IB could not be made, although this may be due to 
the incomplete data available. A number of individual tombs from the 
cemeteries E and U at Abydos, as well as tombs from el-Amrah, can 
be attributed to Naqada IA because of the presence of simple Black- 
Topped cups and beakers belonging to Petrie's types B 18 d, B 21 b 
and B 22 b/d/f. Also frequentiy occurring in tombs attributed to Naqada 
IA is White Cross-Lined pottery, the types of which are mainly sim- 
ple convex round based bowls and plates (C 10 e/l/n), although restricted 
cups (C 64 b/n) and slender restricted jars (C 76 h) are also repre- 
sented. Restricted shapes and lip rims are however very exceptional. 

Naqada IB 

Besides the tombs already mentioned from Naqada Main Cemetery, a 
small group of Naqada IB tombs occurs south of the centre of ceme- 
tery 2600—2700 at Matmar. In the ceramic assemblage, there is a noted 
increase of Red-Polished pottery. The type variation increases both for 
the Black-Topped and White Cross-Lined class. Besides the simple 
Black-Topped cups and beakers with straight wall already mentioned 
for Naqada IA, types with inflected wall, some of them large in size 
(B 18 b/c, B 21 c/d2, B 22 j, B 25 b/c, B 26 b), become characteristic. 
Additional types for the White Cross-Lined are oval plates (C 4 h, 
C 5 d/m, F 1 1 a) and exceptionally fiat based bowls with concave wall 
(C 35). The frequency of slender restricted jars increases. The Red- 
Polished types consist mainly of small bowls and plates (P 1 a, 
P 11 a, P 17), as well as slender jars (C 75 b, C 76 d, C 76 w), the 

Hendrickx (n. 39), 41. 
Friedman (n. 66). 


shapes of which also occur among the Black-Topped and White Cross- 
Lined pottery. 

JVaqada IC 

Groups of Naqada IC tombs can be spatially identified at several ceme- 
teries (Matmar, Salmany, Naqada and Armant). Both cups and beakers 
with straight and flaring walls occur, but the latter group shows increas- 
ing variability. Among the tall beakers, the importance of high, slen- 
der types increases (B 27 a, B 27 f, B 35 a) and types with explicit 
concave upper part (B 26 a-c). Hardly ever attested before are restricted 
regularly curved jars with large aperture (B 55 b, B 57 a/b, B 58 b/c, 
B 62 b/d, B 77 a), or similar shouldered jars (B 71 b, B 74 a/b, B 
79 a). Also new are flat based bottles (B 92 a/b), which also occur 
among the Red-Polished pottery (P 56 a/b). As before, the majority 
of the Red-Polished pottery however reflects shapes known for the 

The majority of the White-Cross lined bowls and plates have, by 
this stage, concave walls and a flat base (C 24, C 26 1, C 27 n, C 30 
h) or inflected walls with concave upper part (C 53-54). Although the 
decoration still primarily consists of geometric motifs (probably imita- 
tion of basketry), figurative representations become more frequent. The 
limited amount of Rough pottery which occurs from now on consists 
mainly of small bowls and plates. Footed stone vessels, almost exclusively 
from basalt, 80 and rhomboidal or fishtail flint knives remain exceptional 
and are of the same types as during Naqada IA-B. There is, on the 
other hand a marked increase in palettes, mainly rhomboidal in shape. 

Naqada II 

The definition of the transition between Naqada I and II is most prob- 
lematic. According to Kaiser's general principles, Stufe I should be dom- 
inated by Black-Topped pottery, which is indeed a fact, and Stufe II 

80 Gf. L. M. Mallory-Greenough, J. D. Greenough &J. V. Owen, "The Stone Source 
of Predynastic Basalt Vessels: Mineralogical Evidence for Quarries in Northern Egypt", 
JAS 26 (1999), 1261-1272. 


by Rough pottery. This however, is not so for Stufe Ila at the Armant 
cemetery, where Black-Topped pottery remains dominant and even for 
Stufe lib the prevalence of Rough over Black-Topped pottery remains 
limited. 81 The differences between Stufe Ila and lib, when the domi- 
nant class of pottery changes from Black-Topped to Rough, and espe- 
cially between Stufe lib and lie, with the introduction of Wavy Handled 
pottery and a number of new Decorated types, are much more impor- 
tant than the difference between Stufe Ic and Ila. It is to be noted, 
however, that the Rough pottery does not appear out of the blue at 
a certain moment in the evolution of the Naqada culture. It is more 
than obvious from settlement excavations that the Rough pottery makes 
up the vast majority of pottery since the beginning of the Naqada cul- 
ture and even already during the Badarian," 2 but the Rough ware finds 
its way only slowly to the cemeteries. 

There are certainly enough reasons to make a distinction between a 
first and a second period within the development of the Naqada cul- 
ture, but it seems more logical to draw the line between Stufe Ila and 
lib or perhaps even between Stufe lib and lie. However, for the revi- 
sion of Kaiser's Stufen chronology presented here, the division between 
Stufe I and II has been maintained between Naqada IC and IIA, the 
archaeological characteristics of which are largely identical with Kaiser's 
Stufen Ic and Ila. This is in order not to introduce yet another com- 
pletely different system which would render the use of older literature 

Naqada IIA 

As for the Naqada IC period, spatially distinguished groups of tombs 
can be identified at the cemeteries of Matmar, Salmany, Naqada and 
Armant. The expansion into Nubia has continued and A-group ceme- 
teries contemporaneous with Naqada IIA B can be found south of 
those already identified for Naqada I. Black-Topped pottery remains 
dominant but White Cross-Lined ware disappears largely while Rough 
pottery for the first time makes up a relevant part of the ceramic assem- 
blage. The distinction between Naqada IC and IIA does not only 

Hendrickx (n. 39), 39-40. 

E.g. Brunton (n. 12); Hendrickx et al. (n. 24). 


depend on the representation of the wares. Of great importance is the 
appearance, during Naqada IIA, of a number of pottery types, espe- 
cially small bag-shaped Rough types with pointed base (R 65 b/c, R 
66 a/p, R 69 c/d), which were not yet present during Naqada IC. 
With regards to the Black-Topped pottery, the very early cups and 
beakers with straight wall disappear almost completely. For the remain- 
ing Black-Topped types there is little difference from Naqada IC, 
although the occurrence of modelled rims is new (e.g. B 35 b, B 37 
b, B 38 c, B 50) as well as large flat based, regularly curved jars with 
strongly marked rim (B 53 a/b). Once again, the shapes of the Red- 
Polished pottery are related to those of the Black-Topped class. 
Rhomboidal palettes still occur regularly but the number of fish shaped 
palettes increases. 

Naqada IIB 

Spatially distinguished groups of tombs can once again be recognised 
at the cemeteries of Matmar, Salmany, Naqada and Armant, already in 
use from at least the Naqada IC period. The quantitative importance 
of Black-Topped pottery starts to diminish, although it is still the best 
represented class. Beakers now occur only with inflected walls and 
mainly exclusively with slender shapes (groups B 25, B 27, B 35) and 
often with modelled rim. The majority of the Black-Topped types con- 
sists of regularly curved jars, a limited number of which has pointed 
bases (groups B 41, B 44) but the majority is flat based (groups B 
56-58, B 62-68). The diversity of Red-Polished types increases and 
is not as closely linked to the Black-Topped types as was previously 
the case. 

The presence of Rough pottery has more than doubled. This is 
reflected in a more important diversity of cups and bowls (R 1-36), 
the previously occurring small bag shaped jars (R 62~69) with pointed 
base and also large storage jars (R 80-86) which will become very 
characteristic for the Naqada IIC-IID2 period. White Cross-Lined pot- 
tery has disappeared completely and is replaced (?) by a limited num- 
ber of Decorated types, the decoration of which is limited to spirals, 
undulating lines and "scales". That the limited number of Late types 
is indeed made from marl clay cannot be confirmed beyond doubt 
given the confusion within this class of pottery, where no clear dis- 
tinction has been made between Nile silt and marl clay pottery. 


There are some indications for regional diversity during Naqada IIB. 
Friedman observed a strong presence of Black-Topped ware for Stufe 
lib at Naqa ed-Deir, 83 and a similar phenomenon was also noted by 
Kaiser for el-Mahasna. 84 For Naqada, Payne noted differences with Stufe 
lib as described by Kaiser. 83 All of this seems to indicate that the tran- 
sition between Naqada IIA/B and Naqada IIC did not happen in the 
same manner and/or at the same moment for the whole of UE. 

Naqada IIC 

Cemeteries and settlements of this period are to be found over a larger 
area than those of the previous period. Besides at the cemeteries already 
mentioned for the previous periods, Naqada IIC was also spatially 
identified at Badari, Hammamiya, Naqa ed-Deir and the Hierakonpolis 
Fort Cemetery. Furthermore, Naqada IIC occurs at cemeteries in the 
neighbourhood of the Fayum (Gerza, 86 Haraga, 8 ' and Abusir el-Meleq), 88 
and in the eastern Nile delta the important cemetery at Minshat Abu 
Omar may have started towards the end of Naqada IIC but probably 
rather during Naqada IID. 89 In Nubia, A-group cemeteries with strong 
Egyptian influences can be found. 

With the Naqada IIC period, major changes occur in the pottery 
assemblage. The importance of the Black-Topped class drops dramat- 
ically and at the same time there is a marked increase of Rough pot- 
tery. The appearance of Wavy-Handled pottery is also very important. 
Although a limited number of the already known types still occurs, 
Black-Topped pottery is now dominated by shouldered jars with a small 
base and modelled rim (B 38 a/c, B 53 a— c) and similar shapes with 
pointed base (B 39 a— b). The Red-Polished pottery already in use dur- 
ing the previous period continues to be used. The most important 
difference is a strong increase of regularly curved jars with a small base 
(P 40 group). The diversity within the Rough class increases, mainly 

83 Friedman (n. 66), 74-5. 

84 Kaiser (n. 53), 74. 

85 Payne (n. 39), 81. 

86 W. M. F. Petrie, The Labyrinth, Gerzeh and Mazguneh (London, 1912). 

87 R. Engelbach & B. Gunn, Harageh (London, 1923). 

88 Scharff (n. 48). 

89 W. Kaiser, "Zum Friedhof der Naqada-Kultur von Minshat Abu Omar", ASAE 
71 (1987), 119-126. 


because of the differentiation within previously occurring groups. The 
large jars (R 81—86), which are a very important part of the Rough 
class, are totally dominated by the type R 81. 

The new Wavy-Handled class is dominated by W 19, but the ear- 
liest examples of this class (W 1-3) also belong to this period although 
their provenance and position is sometimes unclear. 90 The Decorated 
pottery is dominated by relatively small fiat based regularly curved jars 
(e.g. D 43) and broad round based vessels (e.g. D 61), both with lug 
handles. Besides the decorative patterns already present, there a now 
very characteristic figurative representations (D 40—48). Marl clay pot- 
tery is certainly represented from Naqada IIC onwards. Both the Wavy- 
Handled and the Decorated class are made from this fabric, as is 
the large majority of the Late pottery. The latter is primarily used for 
cups and bowls, as well as a number of medium sized shouldered jars 
(L 53 a— b). The types of stone vessels that had occurred up until now 
have disappeared and are replaced by shapes inspired by the Decorated 
pottery. Rhomboidal palettes become rare, and in their place we find 
fish shapes and palettes with antithetic bird's head. 

Naqada IID1 

Cemeteries and settlements are now probably also found in the delta 
(Minshat Abu Omar, Kafr Hassan Daoud) and the gradual replace- 
ment of the Maadi-Buto culture can be observed at Buto (cf. below). 
Black-Topped pottery becomes rare and only a number of regularly 
curved and shouldered jars occur (B 38 c, B 39 a— b, B 53 a— b), all of 
them with modelled rim. For the Red-Polished pottery there is hardly 
any difference from Naqada IIC, both in frequency and typology. A 
similar observation can be made for the Late class, despite its slowly 
increasing importance. Rough pottery continues to represent over half 
of the assemblage. The most important changes are to be found in the 
group of large storage jars where type R 81 is gradually replaced by 
the types R 84 and especially R 85 h and R 86 d, meaning that vessels 
with a small or pointed base are preferred to round based examples. 
The Wavy-Handled types become smaller and more slender (W 25, 
W 44) compared to the types characteristic for Naqada IIC. The two 

Kaiser (n. 48), 87-109. 


Decorated types characteristic for Naqada IIC (D 43, D 61) continue 
to be used but figurative representations occur less frequently. The stone 
vessel types remain identical to those of Naqada IIC, while rhombic 
palettes have almost completely disappeared. 

Naqada IID2 

The definition of Kaiser's Stiffen IId2 and Illal causes a problem because 
they share the same Wavy-Handled types and differ mainly through 
the presence or absence of Black-Topped types, the number of which 
is at any rate minimal, and through their Decorated types, which also 
only account for a very small part of the assemblage. Another char- 
acteristic should be the transition from R 84-86 to L 30 b,c but the 
latter types are in reality very similar to R 84 and R 84 c and the 
difference consists mainly of a less well cared for product. It is there- 
fore to be feared that the attribution of a vessel to one of these types 
by excavators other than Petrie may have been rather arbitrary. Also 
the spatial distribution at Armant Cemetery 1400—1500 easily allows 
for a different clustering of graves, by which the group defined by 
Kaiser as Stufe IId2, no longer exists. 91 Furthermore, the Wavy Handled 
types always seem to display the fastest evolution of shape, and it would 
be very strange if this would not have been so during Stufe IId2-IIIal. 
For all these reasons the archaeological description of the Stiffen IId2 
and Illal cannot be maintained in the way in which it was defined 
by Kaiser and the description given here for Naqada IID2 therefore 
differs to some extent from Kaiser's Stufe IId2. The distribution of ceme- 
teries and settlements remains the same as for Naqada IID1. 

Wavy-Handled and Late pottery increases in importance while all 
other classes decrease (lightly). The types represented for Black-Topped 
and Red-Polished pottery nevertheless remain almost identical. This is 
not the case for Late pottery, where not only the previously known 
cups, bowls and jars (L 53 group) show a greater diversity but fur- 
thermore a new type of large, round based shouldered jars (especially 
L 36 n) occurs. The latter will become increasingly popular from Naqada 
IIIA1 onwards. The large majority of the Rough types present during 
Naqada IID 1 remains so, but with less diversity. A few new types occur, 

Hendrickx (n. 69), 274-276, pi. 87-89. 


which are mainly imitations of Late types (e.g. R 57 b-c). Among the 
large storage jars, the decline of type R 81 already noted for Naqada 
LID 1 continues, for the benefit of R 84 (= L 30 b) and R 86 p (= L 
30 — ). There is a marked change for the Decorated ware. Of the two 
shapes typical from Naqada IIC onwards, the small flat based regu- 
larly curved jars (e.g. D 43) hardly ever occur any more, while the 
broad round based vessels (e.g. D 61) still do. Figurative representa- 
tions are almost completely missing. For the Wavy-Handle class, the 
tendency towards smaller vessels and narrow shapes continues (groups 
W 43 and W 47), while the handles loose their functionality. 

Naqada III 

The transition from Kaiser's Stufe II to Stufe III is not without prob- 
lems. The difference between them is made up by the Late class, which 
takes over from the Rough class as numerically the most important 
group. However, Kaiser's view of the spatial distribution of the Rough 
and Late pottery at Armant 92 does not take into account the fact that 
an important number of the Late types are in reality made in the 
Rough fabric (especially the types belonging to the L 30 series), although 
he is well aware of the problem. 93 Counting these with the Rough class, 
shows that at Armant no part of the cemetery is dominated by marl 
clay pottery. However, this does not mean that groups of graves dom- 
inated by marl clay pottery do not occur during the Naqada culture. 
On the contrary, large groups of graves at Elkab and Hierakonpolis 
for instance, and even entire cemeteries such as those of Tarkhan, Tura 
and Abu Roash are completely dominated by marl clay pottery. Only 
the transition in dominance from Rough to Late pottery should be 
placed more recently than suggested by Kaiser. This raises the ques- 
tion of whether a transition from Naqada II to III should be situated 
at this moment. As we are dealing with a gradual evolution and not 
with sudden changes, this question is only of limited importance. A 
change of main period is not supported by the characteristics of the 
funerary equipment of the tombs. However, as it is certain that at least 
from the beginning of the Naqada III period that the Naqada culture 

Kaiser (n. 53), PI. 15 B-C. 
Kaiser (n. 53), 76, note 9. 


has spread over the whole of Egypt and as the earliest writing and his- 
torical documents occur during the Naqada IIIA1 period (cf. below), 
it seems possible nevertheless to maintain the division between Naqada 
II and III, albeit not for the original reasons. Naqada III sites occur 
throughout Egypt and Nubia and their distribution can no longer be 
used as a possible element for chronological information. 

Naqada IIIA1 

The definition of Kaiser's Stiffen Illal causes a particular problem. The 
number of tombs at Armant for Stufe Illal is very limited, 94 and the 
Wavy-Handled types found in them are very similar to those of Stufe 
IId2. As mentioned already there is no domination by marl fabric pot- 
tery that would make a difference from Stufe IId2. It therefore seems 
impossible to retain Stufe Illal in the way defined by Kaiser and the 
description given here for Naqada IIIA1 will differ strongly. The major- 
ity of Stufe Illal tombs is included in Naqada IID2, while part of Stufe 
IIIa2 is considered a separate entity and redefined as Naqada IIIA1 
(cf. below). 

Black-Topped pottery no longer occurs and the importance of Red- 
Polished pottery has strongly diminished. For most types of the latter 
class, parallels can be found among the Late pottery and it is very pos- 
sible that confusion has occurred. Unfortunately it has not yet been 
possible to check whether the vessels identified as Red-Polished are 
indeed made from Nile silt and not from the marl fabric characteris- 
tic of the Late class. The large Rough storage jars are now dominated 
by slender types (L 30 g/k, L 31 a) which are, however, taller than 
the Naqada IID types. Among the round based shouldered marl clay 
storage jars the types L 36 n/s are particularly well represented. There 
are no important changes in the remaining part of the Late pottery, 
although the diversity and quantity increases. The Decorated pottery 
characteristic of Naqada IIC— IID2 no longer occurs. The types of ves- 
sels, which are now decorated also, occur without decoration among 
the Late class. The decoration is largely reduced to series of undulat- 
ing lines (groups D 20-21, D 24-25). The Wavy-Handled types again 
become more slender (W 49-50) and the handles that had already lost 

94 Only 4 tombs. It is to be noted that on the distribution plan (Kaiser (n. 53), PI. 
20 C), the symbols for Stufe Illal and IIIa2 have been confused. 


their functionality are, in a number of cases, changed into a continu- 
ous ornamental band (W 51 a, W 56 a, W 56 g). Finally it is to be 
noted that palettes show less diversity in shape and are dominated by 
simple rectangular types. 

JVaqada IIIA2 

Kaiser's definition of Stufe IIIa2 at Armant is based on a very limited 
number of graves and the majority of the types of objects presented 
as characteristic for Stufe IIIa2, 9> are not represented at the Armant 
cemetery. The description of Stufe IIIa2 is therefore largely theoretical. 
When studying the spatial distribution of the Naqada III cemetery of 
Elkab two groups could be distinguished within material characteristic 
for Kaiser's Stufe IIIa2. 9b A similar observation could be made for the 
Hierakonpolis Fort Cemetery, 97 which was also not yet published at the 
time when Kaiser made his study of the relative chronology. Because 
of this and the above mentioned problems with the Stufen IId2 Illal, 
the earliest group distinguished within Kaiser's Stufe IIIa2 was read- 
justed to Naqada IIIA1. The validity of this was later confirmed by 
the horizontal development of the elite tombs at cemetery U at Abydos 
and especially the position of tomb U-j and the types of objects found 
in it (cf below). 

The Naqada IIIA2 period is characterised by a spectacular decrease 
of the Rough class in favour of Wavy-Handled pottery. It is to be 
noted however, that this picture is strongly influenced by the Tarkhan 
Valley cemetery from which the majority of the Naqada IIIA2 tombs 
known comes. The possibility for regional variation, especially between 
UE and LE, can therefore not be excluded. The Nile silt Red-Polished 
pottery has now been completely replaced by marl clay Late ware, the 
variety and quantity of which once again increases. The large round 
based shouldered storage jars are now dominated by somewhat broader 
types (L 36 a/g2/k). 98 The large Rough jars consist mainly of very 
slender pointed jars (L 3 1 a), while the series R 84-86 no longer occurs. 

95 Kaiser (n. 53), PI. 24 B. 

96 Hendrickx (n. 70, Elkab V), 205-16. 

97 See the review by S. Hendrickx, BiOr 47 (1990), col. 643-646. 

98 Naqada IIIA2 tombs have been published both according to Petrie's predynastic 
and protodynastic typology (Petrie (n. 42), 1921, (n. 45), 1953). The most important 


Decorated pottery diminishes in quantity, but for the remaining exam- 
ples both the vessel types and the decoration are identical to Naqada 
IIIA1. It is however to be noted that the amount of Decorated vessels 
would be much higher if the Wavy-Handled jars of type W 62 with 
net pattern decoration would be included. These are most character- 
istic for Naqada IIIA2 and are part of the Wavy-Handled jars with 
their maximum diameter located at the rim, relatively broad base and 
continuous decorative band (W 55, W 58, W 61, W 62)." The stone 
vessels consist mainly of calcite bowls and plates. Animal shaped palettes 
have almost completely disappeared. The vast majority are rectangu- 
lar with incised lines around the edges. 

Naqada IIIB 

Kaiser's Stufe Illb does not occur at the Armant cemetery, on which 
his chronological framework is based, and is therefore mainly an inter- 
pretation of the theoretical evolution of the Wavy-Handled types. More 
recently, Kaiser divided Stufe Illb in two subdivisions, Illbl and IIIb2 
and added three Stiffen, IIIcl, IIIc2 and IIIc3. 10() With the late types of 
the Wavy Handled class as main characteristics, the chronological stages 
distinguished by Kaiser are summarised in Table II. 1.5. 101 The dis- 
tinction made between Stufe Illb 1 and IIIb2 does not seem justified, 
since at Tarkhan, for instance, pottery types characteristic of the Stiffen 

concordances for the marl fabric storage jars are: 60 g = L 36 n, 60 j = L 36 k, 
L 60 m = L 36 a. 

99 For the Naqada IIIA2 Wavy-Handled jars, the most important concordances 
between Petrie's predynastic and protodynastic typologies are: 46 b = W 62 -, 46 d = 
W 58, 46 f = W 58 -. 

100 Kaiser (n. 58). In an earlier stage of research, Kaiser distinguished three peri- 
ods, called Horizonte [W. Kaiser, "Einige Bemerkungen zur agyptischen Friihzeit. III. 
Die Reichseinigung", £45 91 (1964), 92-96; W. Kaiser & G. Dreyer, "Umm el-Qaab. 
Nachuntersuchungen im fruhzeitlichen Konigsfriedhof. 2. Vorbericht", AIDAIK 38 (1982), 
260-269]. For discussion of a number of problems related to the Horizonte, see E. C. M. 
van den Brink, "The Incised Serekh-signs of Dynasties 0—1, Part I: Complete Vessels", 
in: Spencer (n. 39), 148-150. 

101 Kaiser (n. 58) gives no archaeological description for these new Stiffen, neither 
does he discuss the way in which they have been distinguished. Therefore, Table II. 
1.5 is based on personal information kindly supplied by W. Kaiser (Poznan symposium 
1992, letter 30 Oct. 1993). The following correlation with the early kings of Egypt can 
be made: Stufe IIIb2 = Iry-Hor and earlier; Stufe IIIcl = Ka — Narmer; Stufe IIIc2 = 
Hor Aha — Djer; Stufe IIIc3 = Djet/Den until the end of Dyn. 1. 


Table II. 1.5. Relative chronological periods as distinguished by Kaiser (n. 53); 

idem (n. 58), Abb. 1 and Hendrickx (n. 69), idem (n. 39), 36-43, illustrated 

by the types of Wavy Handled/Cylindrical jars 

Kaiser 1957, 1990 


Hendrickx 1989, 1996 




no cylindrical jars 


50 t 


50 b-c / h-t 


50 d 


50 d-g 


48 s-t, 49 d, 50 d 




48 s-t, 49 d/1 






47 r-t, 48 s, 49 d/g 


W 50-51 a, 55, 56 g, 61-62 


W 55, 58, 60-62 




W 49-51, 56 a/g 


W 41, 43 b, 47 g 




W 41, 43 b, 47 g 


W 41-42, 43 b, 47 a/g/m 


W 24, 25 


W 24-25, 27 


W 3, 19 


W 3, 19 


Illbl and IIIb2 are very often present together in the same tomb. 102 
Furthermore, the spatial distribution of the two groups of types shows 
no obvious patterning at Tarkhan. Also, the very obvious spatial dis- 
tribution of the Turah cemetery does not support the idea of a chrono- 
logical difference between the above mentioned types. 103 Therefore, the 
description of Naqada IIIB as presented here is more or less equiva- 
lent to both Stufe Illbl and IIIb2. 

In terms of percentages, the distribution of the pottery classes remains 
almost identical to Naqada IIIA2. There are however a number of 
important changes in the types of vessels. The slender Rough jars with 
pointed base have almost completely disappeared. The other Rough 
types hardly change. For the marl clay storage jars, two size classes 
can be distinguished. The smaller types (L 36 b, L 38 a) occur more 
frequendy than the large ones (L 36 a, L 36 k). A rare but notable 
new type is large jars with applied rope decoration, generally consid- 
ered to be wine jars (protodynastic 76). 104 On the one hand, the few 
Decorated vessels remaining are identical to those of the Naqada IIIA1 

102 Hendrickx (n. 39), 58-59. 

103 Hendrickx (n. 39), 59. 

104 p or j n j s jyp e f j arS) see van den Brink "The Incised Serekh-signs of Dynasties 
0—1, Part I: Complete Vessels"; idem, "The Pottery-Incised Serekh-Signs of Dynasties 
0-1. Part II: Fragments and Additional Complete Vessels", Archeo-Ml 1 1 (2001), 23-100. 
His group I — II occurs mainly during Naqada IIIB, group III IV during Naqada IIIC 1 . 


period. Net paintings on Wavy Handled jars, on the other, have almost 
completely disappeared. The Wavy Handled class is now represented 
by cylindrical jars with a small band of incisions instead of an applied 
decorative band (W 71 a, W 80 = 47 r-t, 48 s, 49 d/g). A remark- 
able change is the strong increase in stone vessels. Among them are 
imitations of the cylindrical jars and a large amount of bowls and plates. 

JVaqada IIIC1 

Kaiser's Stufe IIIcl consists of types which are partially characteristic 
for Stufe IIIb2 and partially for Stufe IIIc2. The existence of this kind 
of "transitional period" can of course not be denied, but it is not appro- 
priate to distinguish a period, which has no types of objects charac- 
teristic for that period only. This is especially so since the archaeological 
description of the Stufen is often used for dating individual graves or 
even objects. It therefore seems better to distinguish less periods, and 
admittedly have eventually a slightly less detailed idea of the chrono- 
logical evolution of a cemetery. A limited amount of the types men- 
tioned for Stufe IIIcl is therefore in the present description integrated 
in Naqada IIIB, but the large amount is joined with Kaiser's Stufe IIIc2 
types into Naqada IIICl. 

The Late class now becomes more important than the Wavy-Handled. 
The marl fabric storage jars still consist of large (protodynastic 60 b/d) 
and smaller types (59 group), but there is also a new group of still 
smaller jars (65 group). 105 Very large jars with decorative band (76 
group) already occurred occasionally during Naqada IIIB, but now 
become more frequent. Large Rough jars with pointed base on the 
other hand no longer occur and the same applies to Decorated vessels. 
The Wavy-Handled jars have now evolved into cylindrical jars with 
slightly curved wall and without decorative band (W 90 = 50 d-g). 
The previously occurring stone vessel types continue to be used but an 
increase in squat jars and restricted cups can be observed. 

1<b See S. Hendrickx et al., "Milk, Beer and Bread Technology during the Early 
Dynastic Period", MDAIK 58 (2002), 284-286. 


JVaqada IIIC2 

Naqada IIIC2 largely coincides with Kaiser's Stufe IIIc3. The amount 
of Wavy-Handled cylindrical jars diminishes drastically, but the amount 
of stone vessels shows an almost identical increase. As the latter group 
consists for a considerable part of cylindrical jars, it is obvious that this 
is merely a change of material and not of vessel type. The pottery 
cylindrical jars without decorative band have straight walls and are nar- 
row (protodynastic 50 b— c/h— t) compared to those of Naqada IIIC1. 
The large marl fabric storage jars become far less frequent than pre- 
viously, but the occurrence of the small (protodynastic group 59) and 
even smaller types (group 65) strongly increases. The large wine jars 
with applied band decoration continue to be present. Palettes have 
almost completely disappeared from the tomb equipment. 

Naqada HID 

Naqada HID is less well defined than the other phases of the Naqada 
culture which have been described. This is due to the limited number 
of tombs identified as Naqada HID, the majority of which come from 
Qaw and Abu Roash. 106 This could however be improved as the recent 
excavations at Adai'ma and Helwan have yielded tombs apparently dat- 
ing to Naqada HID. In future, it should be possible to distinguish sub- 
divisions within Naqada HID and probably also regional differences. 
The link with the early Dyn. 3 pottery is probably closer than has been 
suspected before, but this also needs further investigation. It is also 
problematic that an important characteristic of Naqada HID is defined 
in a negative manner, that is to say the absence of cylindrical jars. 
Both of the smaller types of marl fabric storage jars continue to be 
important. The large wine jars with applied band decoration have devel- 
oped into "torpedo" shapes with almost vertical walls and high rims 
(protodynastic 76 s/u/y). In addition, the vessels are made with less 
care. This tendency apparently becomes more explicit as Naqada HID 
advances. Most important is the appearance of new types of vessels, 
which will develop further during the OK. Among them bowls with 

106 The Abu Roash tombs are not integrated in Table II. 1.4a— b because they are 
not published according to Petrie's typologies. 


internal rim, 107 early Maidum bowls and beer jars. 108 Among the stone 
cylindrical jars, nearly all of them in calcite, types without decorative 
band are now more frequent than those with. New is the frequent 
occurrence of dummy calcite cylindrical jars. 

Correlation between Naqada III and Dyns. 0-2 

Before discussing the actual correlation between Naqada III and Dyns. 
0-2, the use of the concepts "Dyn. 0" and "Dyn. 00" should be dis- 
cussed. The first notion was already used on occasion by Petrie and 
Quibell but has only far more recently gained wide acceptance after 
its reintroduction by Kaiser. 109 Dyn. has however been used with 
different meanings and the only consistency is the inclusion of Iry-Hor 
and Ka. Perhaps its use can be defended for referring to the line of 
kings from Abydos identified through inscriptions and tombs at the 
cemeteries U and B. Dyn. 00 on the other hand has only occasionally 
been used, partially overlapping the meaning of Dyn. 0. Unless it's 
significance would be clearly defined and generally accepted, it seems 
better to avoid the use of Dyn. 00. A number of elite tombs can be 
connected by inscriptions or seal impressions with the Late Predynastic — 
Early Dynastic kings. 110 The relative chronological position of these 
tombs was in some cases defined when studying the spatial distribution 
within the Naqada III cemeteries of LE (Tarkhan, Turah). However, 
the royal tombs from Abydos and the elite mastabas from Saqqara are 
mainly attributed to a particular Naqada period by the characteristic 
objects they contained (Table II. 1.6). It is furthermore possible to inte- 
grate the spatial distribution of the elite tombs from cemetery U at 
Abydos, which can be linked to the local late predynastic kings. 111 

107 D. Raue, "Agyptische und nubische Keramik der 1 .— 4. Dynastie", in: W. Kaiser 
et al., "Stadt und Tempel von Elephantine 25V26./27. Grabungsbericht", MDAIK 55 
(1999), 173-189. 

108 Hendrickx et al. (n. 105). 

109 Kaiser (n. 14), 71. 

110 Hendrickx (n. 39), 59-61. 

111 G. Dreyer, Umm el-Qaab: Nachuntersuchungen im fruhzeitlichen Konigsfriedhof. 
3./4. Vorbericht. MDAIK46 (1990), 61-62; idem, Umm el-Qaab. Nachuntersuchungen 
im fruhzeidichen Konigsfriedhof. 5./6. Vorbericht. MDAIK 49 (1993), 36-37; idem, 
Umm el-Qaab I. Das pradynastische Konigsgrab U-j und seine jriihen Schriftzeugnisse (Mainz, 
1998), 18—19; G. Dreyer et al., Umm el-Qaab. Nachuntersuchungen im fruhzeitlichen 
Konigsfriedhof 7./8. Vorbericht. MDAIK 52 (1996), 29-30; R. Gundlach, Der Pharao 


Table II. 1.6. Correlation between Naqada IIIA1 HID and Dyns. 0—1 










S 3120, 



S 3500, 














S 3038, 
S 3338 








S 3035, 

S 3506, 

S 3507, 

S X 





S 3036 






S 3504 






S 3471, 





Hor-Aha / Djer 








B 10/15/19 

S 3357 










B 17/18 (?) 


414, 415, 1100 











1627, 1651 


B 7/9 






B 1/2 






U-t, U-x, U-y, 



54, 64, 89 



U-g, U-h, U-s, 
U-u, U-v 




Scorpion I 







U-a, U-k, U-o, 
U-r, U-qq 

In the southern part of cemetery U at Abydos, a group of mud brick 
lined tombs clearly dates to Naqada IIIA1. It is even possible within 
this group to recognise a west-east evolution from older tombs (U-a, 
U-o, U-qq) to more recent tombs (U-r, U-k, U-j, U-i). 112 The latter 
are continued further to the east by Naqada IIIA2 tombs (U-f (?), 
U-g, U-h). Starting from tomb U-s, which still dates to Naqada IIIA2, 
a series of tombs sets of, linking cemetery U to cemetery B, where the 
tombs of Iry-Hor up to Hor- c Aha are located. Unfortunately the evi- 
dence published for these tombs in the preliminary reports is limited. 
Eventually tombs U-u and U-v still date to Naqada IIIA2. U-t prob- 
ably dates to Naqada IIIB, and because of their position in the ceme- 
tery, this would also have been the case for U-y, U-z and U-x. The 
earliest tombs of cemetery B, those of Irj-Hor (Bl— 2) and Ka (B7/9) 
can be dated to Naqada IIIB on more reliable basis. 

und sein Stoat. Die Grundkgung der agyptischen Konigsideologie im 4. und 3. Jahrtausend (Darmstadt, 
1998), 54, Abb. 54. 

112 Dreyer et al. (n. Ill, 1996), 29-30. 


The pottery found in the tombs of Narmer (B 17—18), Hor- c Aha 
and Djer is characteristic for Naqada IIIC1, making this apparentiy a 
relatively short period. This seems to be confirmed by the limited num- 
ber of Naqada IIIC 1 tombs at Tarkhan compared to those of Naqada 
IIIB and also IIIC2. 113 The tombs dating to the reigns of "Serpent", 
Den and c Adj-ib can be attributed to Naqada IIIC2. The position of 
the tomb of Semer-khet, who apparentiy only reigned for a short period, 
remains uncertain because hardly any objects from his tomb are known, 114 
and no mastabas dating to his reign have been identified at Saqqara. 
The tombs dating to the time of Qa- C a fit well within Naqada HID as 
described above, but although there are certainly similarities with the 
tombs of late Dyn. 2, there are also differences with the limited amount 
of pottery known for the tombs of Per-ibsen 115 and Kha-sekhemwy 116 
at Abydos. As mentioned before, it is at present impossible to make 
well defined subdivisions within Naqada HID but this will certainly be 
possible in the future. 

Radiocarbon Chronology 

Radiocarbon dates for Naqada cemeteries are limited in number, and 
the majority of them have been made a long time ago, resulting in 
questionable dates and important deviations. Furthermore, the calibra- 
tion curves for the (second half of the) 4th millennium BC show impor- 
tant fluctuations with long possible data ranges as a consequence. It is 
generally considered a "bad period" for radiocarbon dating. 117 It is 
therefore impossible to link the phases of the Naqada culture distin- 
guished to an absolute chronology. The limited number of dates avail- 
able, at any rate, does not allow for the distinguishing of chronological 

113 Naqada IIIA2: 488 tombs; IIIB: 306 tombs; IIIC 1 : 73 tombs; IIIC2: 206 tombs. 

114 G. Dreyer et al., "Umm el-Qaab. Nachuntersuchungen im fruhzeitlichen Konigs- 
friedhof 11./12. Vorbericht", MDAIK 56 (2000), 120-122. 

115 Petrie, Abydos I, pi. VII. 

116 Petrie, Abydos I, pi. VII; E.-M. Engel, "Abydos. Umm el-qa'ab, Grab des 
Chasechemui. Deutsches Archaologisches Institut Kairo", BCE 20 (1997), 25-28; idem, 
Abydos, Umm el-Qa'ab, Grab des Chasechemui", BCE 21 (2000), 50-58. 

117 Cf. B. Midant-Reynes & P. Sabatier, "Prehistoire egyptienne et radiocarbone", 
Archeo-Nil 9 (1999), 83-107. 


phases. 118 By also integrating the dates from settlement sites it is nev- 
ertheless possible to confirm the outline defined by Fekri Hassan. 119 

Radiocarbon dates from historically dated royal and elite tombs at 
Abydos and Saqqara have been used in an attempt to date the reigns 
of individual kings accurately. 120 This however is not without problems 
as is shown by the more recently published dates from Abydos, 121 indi- 
cating that Scorpion I (tomb U-j) and Hor-Aha would be more or less 
contemporaneous. Because of the characteristics of the funerary equip- 
ment and the location within the cemetery, this is however to be 
excluded. Eventually the use of old wood could account for this type 
of aberration, something, which might quite easily have occurred in 
the well-organised elite cemeteries. The correlation presented here (Table 
II. 1.7) between the Naqada periods, the Dyns. 0-2 kings and the avail- 
able radiocarbon dates is therefore to be considered preliminary and 
approximate only. A major problem concerning the interpretation of 
radiocarbon dates is the discrepancy between the historical chronology 
and the radiocarbon chronology. The radiocarbon dates are consis- 
tendy older with by least 100 years, 122 resulting in a very long time 
span for Dyn. 2. As this dynasty is unfortunately poorly known, this 
question has to remain open for the time being. Also, the correlation 

118 Contra S. H. Savage, "AMS Radiocarbon Dates from the Predynastic Egyptian 
Cemetery, N.7000, at Naga-ed-Der", JAS 25 (1998), 235-249, cf. A. R. Miljard & 
T. A. H. Wilkinson, "Comment on 'AMS Radiocarbon Dates from the Predynastic 
Egyptian Cemetery, N7000, at Naga-ed-Der by S. H. Savage'", JAS 26 (1999), 339-341. 

119 F. A. Hassan, "Radiocarbon Chronology of Archaic Egypt", JMES 39 (1980), 
203-207; idem, "Radiocarbon Chronology of Predynastic Naqada Settlements, Upper 
Egypt", Current Anthropology 25 (1984), 681-683; idem, "Radiocarbon Chronology of 
Neolithic and Predynastic sites in Upper Egypt en the Delta", AAR 3 (1985), 95-116; 
F. A. Hassan & S. W. Robinson, "High-precision Radiocarbon Chronometry of Ancient 
Egypt, and Comparisons with Nubia, Palestine and Mesopotamia", Antiquity 61 (1987), 
119—135. See also Hendrickx (n. 7), 13-81; Midant-Reynes & Sabatier, Archeo-Ml 9 
(1999), 83-107. 

120 Hassan, JJVES 39 (1980), 203-207. 

121 R. M. Boehmer, G. Dreyer & B. Kromer, "Einige Friihzeitliche 14C-Datierungen 
aus Abydos und Uruk", MDAIK49 (1993), 63-68; J. Gorsdorf, G. Dreyer & U Hartung, 
"14C Dating Results of the Archaic Royal Necropolis Umm el-Qaab at Abydos", 
MDAIK54 (1998), 169-175; idem, "New 14C Dating of the Archaic Royal Necropolis 
Umm el-Qaab at Abydos (Egypt)", Radiocarbon 40 (1998), 641-647; Dreyer (n. Ill, 
Umm el-Qaab I), 17-19. 

122 Gorsdorf, Dreyer & Hartung, MDAIK 54 (1998), 175. See also H. Haas et al., 
"Radiocarbon Chronology and the Historical Calendar in Egypt", in: O. Aurenche, 

J. Evin & F. Hours, eds., Chronologies in the Near East. Relative Chronologies and Absolute 
Chronologie 16,000-4,000 B.P. (Oxford, 1987), 585-606. 



Table II. 1.7. Absolute chronology 

cal. BC 

Naqada HID 
Naqada IIIC2 
Naqada IIIC1 
Naqada IIIB 
Naqada IIIA2 
Naqada IIIA1 
Naqada IIC-IID2 
Naqada IA IIB 

from ca. 2920 onwards 
ca. 3000-2920 
ca. 3150-3100 

ca. 3350-3150 

ca. 3600-3350 

ca. 4000/3900-3600 

[Semerkhet]/Qa-a - Dyn. 2 

Djet - Adjib 

Narmer - Djer 

U-t, Iry-Hor - Ka 


U-a,k,o,r,qq - Scorpion I 

Table II. 1.8. Concordance between the Buto stratigraphy and the relative 
chronologies of the Naqada culture and the Southern Levant (after E. C. Kohler, 
Tell el-Fara'in — Buto III. Die Keramik von der spaten Naqada-Kultur bis zum friihen Alten 
Reich (Schichten III bis VI) (Mainz, 1998); Faltings, in: Interrelations, 165—170; Levy 
& van den Brink, in: Interrelations, 3-38) 



Naqada period 

Southern Levant 

Buto V 

Naqada IIIC2-IIID 


Buto Illf-IV 

Naqada IIIB IIIC 1 

EB IB (late) 

Buto Hid e 

Naqada IIIA2 

EB IB (middle) 

Buto Illb-c 

Naqada IIIA1 

Buto Ilia 

Naqada (IID1)-IID2 

EB IB (early) 

Buto lib 

Naqada IIC-IID1 

Buto Ha 

Wadi Digla II 

Naqada IIA-B 


Buto lb 

Wadi Digla I 

Naqada IC 

Buto la 

(Badari) - Naqada IA-B 

Late Chalcolithic 

between radiocarbon dates from the southern Levant and Egypt is 
equally problematic. 123 

Maadi-Buto Culture 


The importance of this culture has only been realised during the last 
decade. The Maadi-Buto culture, dating mainly to the 4th millennium 
BC, is no longer regarded as a regional culture in the neighbourhood 

123 E. Braun, "Proto, Early Dynastic Egypt and Early Bronze I— II of the Southern 
Levant: Uneasy 14C Correlations", Radiocarbon 43 (2001), 1279-1295. 

124 Only a short overview is presented here. More detailed archaeological descrip- 
tions can be found in the literature mentioned. 


of Maadi, today a modern suburb of Cairo, because it was apparently 
present over a large part or perhaps even the whole of LE, especially 
during Naqada IIC/D. 125 This, however, does not necessarily implicate 
a political unity of LE. 12b The connection between the different chrono- 
logical phases, early, middle and late Maadi-Buto, recognised at Maadi 
and Buto between themselves and in relation with the relative chronol- 
ogy of the Naqada culture, 127 poses considerable problems and the 
earlier phases in particular have been revised several times over recent 
years. At present they can nevertheless be summarised with some 
precision. 128 

Of great importance is the presence in the lowest layer of occupa- 
tion at Buto, Stratum Ia/b, of a large quantity of locally made south 
Levantine Chalcolithic-style pottery, known from the Beersheba culture 
in the southern Levant. 129 Stratum la, containing about one-third of 
locally made Chalcolithic pottery, is to be dated before 3700 cal BC, 130 
contemporaneous with Naqada IA/B. The remaining part consists of 
LE Neolithic pottery. The amount of south Levantine Chalcolithic-style 
pottery diminishes in Stratum lb, which could be contemporaneous 
with Naqada IC, although this is mainly based on the intermediate 
position of this stratum between Buto la and Ila for which more reli- 
able links with the Naqada culture are available. Buto II represents the 
"classical" stage of the Maadi-Buto culture and the two layers distin- 
guished, Buto Ila and b, can be dated to respectively Naqada IIA B 
and Naqada IIC-IID1. Buto Stratum Ilia, corresponding mainly to 
Naqada LID 2, represents a transitional phase between the LE Maadi- 
Buto culture and the Naqada culture which by that time has started 
spreading northward. The influence of the Naqada culture in the Delta 
becomes far more important during Buto Stratum Illb c, correspond- 
ing roughly to Naqada IIIA1. During Buto Stratum Hid f hardly any 
elements of the Maadi-Buto culture remain. (See also page 487, first 

125 Levy & van den Brink, in: Interrelations, 11-13. 

126 See however Levy & van den Brink, in: Interrelations, 8. 
12/ Levy & van den Brink, in: Interrelations, 13, Table 1.4. 

128 Fairings, in: Interrelations, 165-170; Levy & van den Brink, in: Interrelations, 19, 
Table 1.8. 

129 Fairings, in: Stationen, 35-45; idem (n. 6, Proceedings); idem, in: Interrelations 
165—170. See also C. Commenge & D. Alon, "Competitive Involution and Expanded 
Horizons: Exploring the Nature of Interaction between Northern Negev and Lower 
Egypt (c. 4500-3600 BCE)", in: Interrelations, 139-153. 

130 Fairings, in: Interrelations, 168. 


Jochem Kahl 


Contemporaneous king-lists intended as historical records are not at 
our disposal for reconstructing the chronology of the earliest dynasties 
nor for establishing the order of the kings. That such lists existed is 
shown by the Royal Annals. 2 The only contemporaneous sources are 
of a different nature, viz. a) kings' names inscribed on ritual or festival 
vessels or on administrative seals (and in one case, on a non-royal 
statue), useful in reconstructing the succession of kings; b) administra- 
tive labels citing eponymous events of specific years and stone vessels 
inscribed with festival notes which aid in determining the length of 
reigns. Later sources used for this study are restricted to the OK. It 
must be borne in mind that all sources, whether contemporaneous or 
later, may contain scribal errors,'' and that some, such as the Royal 
Annals, were subject to ideological influence. 

Kings before Nar-mer 

The earliest writing from Egypt preserves the names of several rulers 
who preceded Nar-mer, here considered the first king of Dyn. I. 4 Two 

1 I would like to thank Eva-Maria Engel and Barbara KneiBler for information and 
for help in preparing the documents. 

2 Wilkinson, Annals; M. Baud, "Les frontieres des quatre premieres dynasties. Annales 
royales et historiographie egyptienne," BSFE 149 (2000), 32-46; idem, "Menes", 109 
147. — Schafer, Annalen, remains the fundamental publication for the Palermo Stone, 
the largest fragment of annals that has survived. 

3 E.g. the labels Petrie, RT I, pi. 15: 16-17 where the meaning of the eponymous 
event was misunderstood — see G. Dreyer et al., "Umm el-Qaab, Nachuntersuchungen 
im fruhzeitlichen Konigsfriedhof, 11./12. Vorbericht," MDAIK 56 (2000), 116 n. b; or, 
for the reign of Den, an entry in the Annals (Cairo Fragment 5, recto, lower part, 5) 
mentioning the planning (?) of a building which must have been erected under Djer, 
cf. Wilkinson, Annals, 246-247. 

4 Recent scholarship favours Narmer over c Aha for the role of Menes, first king of 


of them are documented in the necropolis of This at Umm el-Qaab, 
Abydos: Iry-Hor and Sekhen/Ka.'' Archaeological evidence makes 
Sekhen/Ka the predecessor of Nar-mer and thus a successor of Iry- 
Hor. 6 The presence of the element Hor in the name Iry-Hor can be 
cited in support of this interpretation, since the names of some other 
"kings" attested before Nar-mer also include the god's name: Ny-Hor, 
Hat-Hor, and Pe-Hor. These "kings," as well as some others, are pre- 
sumed to have been local rulers or rulers who opposed the Thinite 
elite. Information for determining the chronological relationship of these 
rulers is provided solely by archaeological evidence.' The following 
names are attested: 

Ny-Hor, at Tura 8 

Hat-Hor, at Tarkhan 9 

"Trio" (three circles surmounting vertical strokes), from the eastern 

Delta, 10 and perhaps also at Tura 11 

Pe-Hor (alternatively read Iry-Hor and thus assignable to him), at 

Qustul 12 

Ny-Neit(?), at Helwan 13 

"Crocodile", at Tarkhan 14 

"Bird and vertical sign", at Tarkhan 15 

the First Dynasty; cf. Kitchen, RITA II, 533—534. Support for this interpretation is 
provided by Docs. 1 and 2 (see infra); cf. L. Morenz, "Gegner des Narmer aus dem 
Papyrus-Land: AWand W'-S, GM 189 (2002), 88. Baud, "Menes", 109-110, provides 
a summary of the arguments pro and contra both Nar-mer and c Aha. 

5 Both readings are possible; see Kahl, System, 38-40. 

6 W. Kaiser & G. D. Dreyer, "Umm el-Qaab, Nachuntersuchungen im fruhzeitlichen 
Konigsfriedhof, 2. Vorbericht," MDAIK 38 (1982), 238. 

7 See J. Kahl, "Hieroglyphic Writing during the Fourth Millennium BC: An Analysis 
of Systems," Archeo-Ml 11 (2001), 106, fig. 3, and cf. W. Kaiser & G. Dreyer, MDAIK 
38 (1982), 260-69; T. von der Way, Untersuchungen iur Spdtvor- und Friihgeschichte Unteragyptens 
(Heidelberg: SAGA 8, 1993), 101. 

8 Junker, Turah, 147, fig. 57; Kaiser and Dreyer (n. 6), 260-69. 

9 Kaiser & Dreyer (n. 6), 260-69. 

10 H. G. Fischer, "Varia Aegyptiaca," JARCE 2 (1963), 44-47. 

11 Junker, Turah, 46-47, fig. 57; cf. Fischer (n. 10); von der Way (n. 7), 101. 

12 B. B. Williams, The A-Group Royal Cemetery at Qustul: Cemetery L (Chicago: OINE 
III, 1986), 163, pi. 76. 

13 Identified in an inscription on ajar uncovered by Z. Y. Saad at Helwan in 1949 
or 1950: E. C. Kohler & E. C. M. van den Brink, "Four Jars with Incised &reA;/!-Signs 
from Helwan Recently Retrieved from the Cairo Museum," GM 187 (2002), 65-66, 
76 fig. 1: 2, 77 fig. 2: 2. 

14 G. D. Dreyer, "Horus Krokodil, ein Gegenkonig der Dynastie O," in: The Followers 
of Horns. Studies Dedicated to Michael Allen Hoffmann 1944-1990, R. Friedman & B. Adams, 
eds. (Oxford: ESAP 2, 1992), 259-63. 

15 W. M. F. Petrie, Tarkhan I and Memphis F(1913), pi. 31:71; Dreyer (n. 14), 260. 


"Scorpion", at Hierakonpolis 16 

a ruler with an obscure name, at Buto" 

Not included in this list is a group of signs consisting of two falcons 
on a serekh ("Double Falcon"), known from Abydos, Tura, Beda, and 
the Sinai. 115 Whether it represents a king's name or is symbolic of royal 
authority per se, remains open. Several groups of signs on labels and 
in inscriptions on vessels from Tomb U-j at Umm el-Qaab, as well as 
signs on the Min colossi from Coptos, on the Libya Palette and on 
some other small finds, have been understood as kings' names. 19 But 
this interpretation is problematic. 20 The groups may be place names 
and/or the names of gods instead. 21 

JVar-mer to Qa- C a: The Succession 

Inscriptions preserved in seal impressions and on stone vessels have 
established a highly reliable model for the succession during Dyn. 1. 
Impressions of two different seals associated with the administration of 
the necropolis were found in the royal cemetery of Umm el-Qaab, 
Abydos. Both seals listed kings who were buried there. One dates from 
the time of Den or Adj-ib (Doc. 1); the other is temp. Qa- C a or Hetep- 
sekhemwy (Doc. 2). According to Werner Kaiser, whose interpretation 

16 Quibell, Hierakonpolis I, pi. 25 (bottom), 26c; cf. G. Dreyer, "Ein Siegel der 
fruhzeitlichen Konigsnekropole von Abydos," MDAIK 43 (1987), 41-42; W. Kaiser, 
"Zum Siegel mit frtihen Konigsnamen von Umm el-Qaab," MDAIK 43 (1987), 116—17; 
Helck, Thinitenzeit, 92; W. Kaiser, "Zur Entstehung des gesamtagyptischen Staates," 
MDAIK 46 (1990), 291 n. 23. 

17 Von der Way (n. 7), 99, fig. 22: 6. 

18 For documentation and discussion see E.-M. Engel, "Ein weiterer Beleg fur den 
Doppelfalken auf einem Serech," in press; Junker, Turah, 47 fig. 57; J. Cledat, "Les 
vases de El-Beda," ASAE 13 (1914), 119 figs. 3-4, 120 fig. 6; E. D. Oren, "Sinai," The 
New Encyclopedia of Archaeological Excavations in the Holy Land, E. Stern, A. Levinson Gilboa 
& J. Aviram, eds., vol. 4 (Jerusalem, 1993) 1388; Kaiser & Dreyer (n. 6), 260-269. 

19 G. Dreyer, Umm el-Qaab I. Das prddynastische Konigsgrab U-j and seine frtihen Schriftzeugnisse 
(Mainz: AV 86, 1998), 178-80. 

20 See B. J. Kemp, "The Colossi from the Early Shrine at Coptos in Egypt," Cambridge 
Archaeological Journal 10 (2000), 211-242; J. Kahl, "Das Schlagen des Feindes von Hu: 
Gebel Tjauti Felsinschrift 1," GM 192 (2003), 47-54. J. Baines, "The earliest Egyptian 
writing: development, context, purpose", in: S. D. Houston, ed., The First Writing. Script 
Invention as History and Process (Cambridge, 2004), 150-189. 

21 So J. Kahl, "Die frtihen Schriftzeugnisse aus dem Grab U-j in Umm el-Qaab," 
CdE 78 (2003), 112-135. 


is followed here, the metal cylinder seal used to make the impressions 
subsumed under Doc. 1 was cut during the reign of Den and altered 
after his death. 22 This accounts for the unsymmetrical sequence of 
kings' Horus names (written without serekhs)' ri and the name of the god 
Khent-imenty. The royal names are arranged from left to right: 
Nar-mer — c Aha — Djer — "Serpent" — Den — Meret-neit. (The sequence 
Djer — "Serpent" — is confirmed by Doc. 3, see infra.). Social, rather 
than chronological reasons dictated that Meret-neit follows Den; as 
king's mother, her status was lower than her son's. 24 This interpreta- 
tion is supported by the material (limestone) 2 ' of the stelae for her tomb 
at Umm el-Qaab and by seal impressions found there which are closer 
to those from the time of "Serpent" than to those temp. c Adj-ib. 26 That 
Meret-neit is presumed to have served as regent for her son accounts 
for her burial among the kings at Umm el-Qaab. 27 Accordingly, the 
chronological order should be Nar-mer — c Aha — Djer — "Serpent"- 
Meret-neit — Den. 

Several impressions from sealings of leather bags facilitated the recon- 
struction of a second cylinder seal, Doc. 2. Its design is similar to that 
of Doc. 1. Again, there are no serekhs. The sequence of kings' names 
from Nar-mer to Qa- C a is reversed, with mention of Khent-imenty, per- 
haps as tutelary deity of the necropolis. 28 Meret-neit is omitted, prob- 
ably because of her lower status. 29 The royal names, arranged from left 
to right, are: Qa- C a — Semer-khet — c Adj-ib — Den — "Serpent" — Djer — 
c Aha — Nar-mer. 

Inscriptions on stone vessels corroborate the succession Den — c Adj- 
ib — Semer-khet — Qa- C a (Docs. 4-8) or segments of it (Docs. 9-13, 
15—17). Paleographical analysis shows that kings' names were added 
from reign to reign. Sometimes c Adj-ib's name was erased (Docs. 10—11; 

22 Kaiser (n. 16, 1987), 119. 

23 Dreyer (n. 16), 35, argues that this was intended to designate the rulers in ques- 
tion as deceased. 

24 So both Dreyer (n. 16), 37, and Kaiser (n. 16, 1987), 118 n. 13. 

25 Not one of the hard stones (grano-diorite, granite, or basalt) used for kings' ste- 
lae since the reign of Den; cf. Fischer (n. 10), 41-43. 

26 Kaplony, Inschriften I, 495-496. 

2/ A seal impression from Saqqara, tomb S 3503 may name Djer and Meret-neit; 
cf. W. B. Emery, Tombs II, 169 (2), fig. 226; Kaplony, Inschriften II, 1183 (730); III, 
fig. 730; Helck, Thinitenzeit, 101. 

28 Dreyer et al., "Umm el-Qaab. Nachuntersuchungen im fruhzeitlichen Konigsfriedhof 
7./8. Vorbericht," MDAIK 52 (1996), 73. 

29 So Dreyer et al. (n. 28), 72. 


cf. Doc. 15). Only a few of the inscriptions on these vessels use Horus 
names; in most cases they mention instead the nsw bjt title + 
name. 30 Because other contemporaneous inscriptions give both names 
of a king, it is possible to equate Horus Den with nsw-bjt Khasty, 31 
Horus c Adj-ib with nsw bjt Mer-pi-bia, 32 Horus Semerkhet with nsw 
bjt Iry-netjer 33 and also with a second unreadable nsw bjt 
name, 34 and, finally, Horus Qa- C a with nsw bjt Sen, 35 nsw bjt 
Sehetep 36 and nsw bjt Qa- C a. 3/ 

The three nsw bjt names of Qa- C a can be interpreted as indica- 
tive of chronologically different periods of his reign. According to this 
proposal, Sen is the oldest of the three. On year labels of Qa- C a, 38 Sen 
is associated with the official Henu-ka 39 who is documented under 
Qa- C a's predecessor Semer-khet. 40 The nsw bjt name Sen and the 
official Henu-ka are both mentioned in connection with a "Sixth 
Occasion of Inspection"; another document citing the same event men- 
tions the nsw bjt Sehetep. 41 Therefore, Sehetep will have replaced 

30 For the controversy about whether the element belongs to the name or to 
the title, cf. Wilkinson, Egypt, 206-207. 

31 Label, Umm el-Qaab, Tomb T: Petrie, RT I, 22, 40-42, pis. 11: 14, 15: 16; 
Helck, Thinitenzeit, 101, 123, 159, 169-170, 172-174, 188, 234; seal impression, Umm 
el-Qaab, Tomb T: Petrie, RT II, pi. 19: 151; Kaplony, Inschriften I, 127; 77, 807 (730); 
seal impression, Umm el-Qaab, Cemetery T (?): Kaplony, Inschriften II, 1104 (83); III, 
fig. 83; seal impression, Umm el-Qaab, Tomb T: Petrie, RT II, 25, 49, pi. 7: 5-6; 
Kaplony, Inschriften II, 1142 (364); III, fig. 364; label, Umm el-Qaab: G. Godron, Etudes 
sur le Horus Den et quelques problemes de I'Egypte archaique (Geneva, 1990), pis. 1: 1 — 3: 6; 
seal impression, Saqqara, Tomb S 3506: Emery, Tombs III, 68-69 (18), pi. 79: 18; 
Kaplony, Inschriften II, 1118 (196); III, fig. 196; Helck, Thinitenzeit, 191; seal impression, 
Abu Roash, Tomb M XII: P. Montet, "Tombeaux de la I re et de la IV e dynasties a 
Abou-Roach (deuxieme partie) — Inventaire des objets," Kemi 8 (1946), 205-12, pi. 14; 
Kaplony, Inschriften I, 135 (W); III, fig. 195; Helck, Thinitenzeit, 191. 

32 Seal impression, Umm el-Qaab, Tomb X: Kaplony, Inschriften III, fig. 245. 

33 Seal impression, Umm el-Qaab, Tomb U: Kaplony, Inschriften III, fig. 229. 

34 Label, Umm el-Qaab, Tomb Q: E.-M. Engel, Das Grab des Qa'a in Umm el-Qa'ab: 
Architektur und Inventor (Diss., microfiche, Gottingen 1997), 437, fig. 217: 5; label, Umm 
el-Qaab, Tomb Q Dreyer (n. 28), 73-74, pi. 14d. 

35 Three labels, Umm el-Qaab, Tomb Q Petrie, RT I, pi. 17: 29; Petrie, RT II, 
pi. 8: 3, 12: 6; Petrie, Abydos I, pi. 11: 11. 

36 Label, Umm el-Qaab, Tomb Q Dreyer (n. 28), 74-75, pi. 14e. 

37 Stone vessel, prov. not known: Kaplony, Steingefdsse, 26-32, pis. 4, 20 (32). 

38 Labels from Umm el-Qaab, Tomb Q Engel (n. 34), 455, figs. 221: 1 and 4. 

39 Cf. Kaplony, Inschriften I, 579. 

10 Cf. labels from his reign: Engel, (n. 34), 437, fig. 217: 1; Petrie, RT II, pi. 8: 5. 
41 Two labels, Umm el-Qaab, Tomb Q Engel (n. 34), 455, figs. 221: 1 and 4 (Sen); 
label, Umm el-Qaab, Tomb Q: Dreyer (n. 28), 74-75, pi. 14e (Sehetep). 


Sen in that year. The "Sixth Occasion of Inspection" could not have 
taken place before the king's sixth year. 42 The new nsw bjt name 
Qa- C a for the king is the youngest because it is associated with his sed- 
festivals. 43 

The style and content of inscriptions attesting two enigmatic kings 
(Horus Senefer-ka 44 and Horus "Bird") 45 date them to the time of Qa- 
c a or slightly later. 46 Three explanations are possible: (a) Senefer-ka and 
"Bird" were rivals of Qa- C a. At the beginning of his reign, Qa- C a had 
the "peaceful" name Sen, "the one who fraternizes." The change to 
Sehetep, "the one who pacifies" and to Qa- C a "the one with raised 
arm" reflect political developments, viz. Q_a- C a opposition to and even- 
tual victory over two opponents. This alternative is favoured here, (b) 
The names Senefer-ka and "Bird" are also names of Qa- C a; i.e., he 
also changed his Horus name in the course of his reign. 47 (c) The names 
belong to rulers who reigned after Qa- C a died. The brevity of their 
reigns did not permit either to arrange for Qa- C a's burial nor were any 
seals cut. Seal impressions found in Tomb 0_ at Umm el-Qaab leave 
no doubt that Hetep-sekhemwy buried Qa- C a there. 48 

Evidence for the Lengths of Reigns during Dyn. 1 

During the Early Dynastic Period a regnal year was not numbered but 
identified by one or more specific significant events occurring in its 
course. 49 This is inferred from the existence of different names for the 
same year in contemporaneous sources (labels and stone vessels) and 
confirmed by the information provided by the Royal Annals. Perhaps 
the use of more than one event resulted from the necessity to "name" 
a year when it began, at a time when only scheduled festivals and 

42 Presuming that there was only one "occasion" in any given year. 

43 Stone vessel, Umm el-Qaab, Tomb Q: Petrie, RT I, pi. 9: 8; stone vessel, pri- 
vate collection: Kaplony, Steingefasse, 26-32, pis. 4, 20 (12). 

44 Lauer, Pyramide IV. 1, pi. 17: 86; Emery, Tombs III, pi. 38: 1; Kaplony, Steingefasse, 
33 (13). 

45 Petrie, RT II, pi. 8A: 6; Lauer, Pyramide IV. 1, pi. IV: 17. 

46 For the sake of completeness, mention should be made of a seal impression from 
Tomb S 3505 at Saqqara which preserves traces of an otherwise unknown Horus 
name; see Kaplony, Inschriften I, 147, 149; III, fig. 742. 

4/ Cf. Lauer, Pyramide IV. 1, 15 (86), with reference to the name Senefer-ka. 

48 Dreyer (n. 28), 71. 

49 Cf. Baud, "Menes", 109-147; idem, BSFE 149 (2000), 32-46. 


ritual or cultic events could be selected to identify the year. At year's 
end, events unforeseen at its beginning, such as expeditions or cam- 
paigns, could have been chosen and cited retrospectively.' In general, 
isolated eponymous events are unsuitable for determining the sequence 
of regnal years, by contrast to sed festivals and other recurring events. 
Even if sed festivals were celebrated for the first time before regnal 
year 30, 51 mention of one suggests a point later, rather than earlier 
in a given reign. A sed festival is documented for Den 52 and Adj-ib; 53 
Qa- C a celebrated a second. 54 Other eponymous events which took place 
repeatedly are also suggestive of a long reign. Examples are the "Sixth 
Inspection" mentioned above and a "Second Running of Apis" during 
the reign of Qa- C a. 55 However, it is not known whether these events 
occurred at regular intervals. Thus they provide only a vague idea of 
reign length at best. 

The Royal Annals, which survive on fragments in Palermo, Cairo 
and London, 56 list entries for every year of Semer-khet's reign (cf. infra). 
For other kings, only some years are preserved. 

50 Similarly, Dreyer, MDAIK 56 (2000), 116 n. a. 

51 Hornung & Staehelin, Sedfest, 62-63. 

52 Label, Umm el-Qaab, Tomb T: Petrie RT, I, 21-22, 40-41, pis. 11: 5, 14: 12; 
Helck, Thinitenzeit, 71, 123, 160, 169-70, 215; label, Umm el-Qaab, Tomb T: G. 
Dreyer, MDAIK 46 (1990), 80, pi. 26a. The mention of a second sed festival without 
a king's name on the fragment of a stone vessel from Umm el-Qaab, Tomb T (Dreyer, 
MDAIK 46 (1990), 80, fig. 9 and pi. 26d) may refer to Den; alternatively, the frag- 
ment may be an intrusive find from the tomb of 'Adj-ib, of Semer-khet or — most 
probably — of Qa- C a. 

53 Stone vessel, Step Pyramid, Gallery H: Lauer, Pyramid* IV. 1, pi. Ill: 7; IV.2, 19-20 
(35); Helck, Thinitenzeit, 123-124; stone vessel, Umm el-Qaab, Tomb X, Q U: Lauer, 
Pyramide IV. 1, pi. Ill: 6; IV.2, 20; Petrie, Abydos I, pi. 5 (upper left); idem, RT I, 20-21, 
39-40, pis. 6: 2, 7: 5 and 10, 8: 11; Helck, Thinitenzeit, 123, 192, 228; stone vessel, 
Saqqara, Tomb S 2446: Quibell, Archaic Mastahas (Cairo, 1923), 13, 41, pi. 33: 5; 
Lauer, Pyramide IV. 1, pi. Ill: 4; IV.2, 20. Cf. K. O. Kuraszkiewicz, "Noch einmal zum 
zweiten Sedfest des Adjib," GM 167 (1998), 73-75. 

54 Sed festival: stone vessel from Umm el-Qaab, Tomb Q Petrie, RT I, 20-21, 40, 
pi. 8: 7-7a; second stone vessel from Tomb Q Petrie, RT I, pi. 9: 8; stone vessel from 
the Step Pyramid, outside Galleries H and B: Lauer, Pyramide IV. 1, pi. IV: 4; IV.2, 
24-25 (42); stone vessel in a Swiss private collection: Kaplony, Steingefdsse, 34-38 (16), 
pi. 5: 22. 

Second .sW-festival: stone vessel from Saqqara, Step Pyramid, outside Galleries H 
and B: Lauer, Pyramide IV. 1, pi. IV: 5; IV.2, 25 (43); another stone vessel from the 
Step Pyramid, Gallery B: Lauer, Pyramide IV. I, pi. 8: 41; IV.2, 24 (41); stone vessel in 
a Swiss private collection: Kaplony, Steingefdsse, 26-32, pis. 4, 20 (12). 

55 Label, Umm el-Qaab, Tomb Q Engel (n. 34), 464, fig. 224. 

56 See note 2, above. 



c Aha: highly probable x+2 years (end of his reign)" and probably 

x+5+y years (later years of his reign)* )!i 

Djer: highly probable 10+y years (beginning of his reign) 59 and x+9+y 

years (middle of his reign) 60 

Den: x+6+y years (middle of his reign) 61 and highly probably x+14+y 

years (later years of his reign) 62 x+2 years (end of his reign) 63 

Semer-khet: 9 years (his complete reign) 64 

Qa- C a: 2+y years (beginning of his reign) 6,1 

Reconstructions of the Annals differ widely and must be considered 
highly speculative. Here statements about hypothetical reign lengths are 
disregarded. 66 The possibility cannot be ignored that Nar-mer's reign 
was also recorded in the annals, especially now following on the dis- 
covery of a label at Umm el-Qaab citing one of his years. 67 

Table II. 2.1. The kings from Nar-mer to Qa-'a 

Horus Names 

nsw bjt Names 

Sed Festivals 

Other Rulers 




« Serpent » 









Iry-netjer and an 
unreadable name 




Senefer-ka, "Bird" 



Palermo Stone, recto, II: 1—2; cf. Wilkinson, Annals, 90—91, fig. 1. 

Cairo fragment 5, recto, upper register; cf. Wilkinson, Annals, 238-40, fig. 10. 

Palermo Stone, recto, II: 3-12; cf. Wilkinson, Annals, 90, 92-103, fig. 1. 

Cairo fragment 1, recto, II; cf. Wilkinson, Annals, 186-93, fig. 4. 

Cairo fragment 5, recto, lower register; cf. Wilkinson, Annals, 240—47, fig. 10. 

Cf. Wilkinson, Annals, 103-19, fig. 1; Baud, BSFE 149 (2000), 37. 

Cairo fragment 1, recto, III: 
Cairo fragment 1, recto, III: 
Cairo fragment 1, recto, III: 

1-2; cf. Wilkinson, Annals, 193-94, fig. 4. 
3-11; cf. Wilkinson, Annals, 194-201, fig. 4. 
12-13; cf. Wilkinson, Annals, 201-202, fig. 4. 

For an instructive overview, see Wilkinson, Annals, 256-57. 
Dreyer, MDAIK 54 (1998), 139, fig. 29 and pi. 5c. 


Hetep-sekhemwy to JVetjery-khet. M The Succession 

The sequence of three Dyn. 2 kings is secure: Hetep-sekhemwy, who 
buried Qa- C a at Umm el-Qaab 69 — Ra c -neb — Ny-netjer. The inscription 
on the shoulder of CG 1 (Doc. 18), a statue depicting a kneeling man 
(presumably a priest), lists these Horus names in that order. Additional 
support for the sequence is provided by inscriptions on stone vessels in 
different hands mentioning Hetep-sekhemwy and Ra c -neb (Docs. 19, 
20). 70 In an inscription mentioning the ^a-house of Hetep-sekhemwy on 
another stone vessel from the Step Pyramid (Doc. 21), the name Ny- 
netjer is written over an erased name. This document, along with the 
inscribed fragment of a stone vessel from the tomb of Per-ibsen (Doc. 
22), substantiates the sequence on the statue Doc. 18. (During Ny-net- 
jer's reign, Ra c -neb's name was erased several times; see Doc. 20, 21 
(?), 22.) 71 

nsw bjt Weneg' 2 is attested only by inscriptions on stone vessels 
found in the Step Pyramid and in Tomb S 3014. /S Weneg's exact posi- 
tion, as well as the identification of his Horus name among those known, 
has remained open until now. 74 A long- known inscription from Tomb 
P at Umm el-Q_aab (Doc. 22) provides the key to solving some of the 
problems associated with Weneg. 75 In the inscription the nsw bjt 
name Ny-netjer faces the opposite direction from the name of Ra c -neb 
and that of his palace (Fig. II. 2.1). Ra c -neb's name is partially erased. 

68 Here and below the Horus name Netjery-khet is used to identify the first king of 
Dynasty 3, in preference to Djoser, since the latter is not documented in contempo- 
raneous texts. 

69 See n. 48, above. 

70 For the sequence Hetep-sekhemwy — Ra'-neb, cf. the arguments of H. G. Fischer, 
"An Egyptian Royal Stela of the Second Dynasty," Artibus Asiae 24 (1961), 47-48 with 
n. 11. 

' ' A stone vessel inscribed with the names of Qa-'a and Ra'-neb (Kaplony, Steingefasse, 
34-38 (16), pi. 5: 22) does not help to reconstruct the order of these kings. 

72 For the reading, cf. B. Grdseloff, "Notes d'epigraphie archaique," ASAE 44 (1944), 

73 Lauer, Pyramide IV. 1, pis. V: 4, 19: 105, 20: 101-103 and 106-107; IV.2, 50-53. 

74 Helck, Thinitenzeit, 103, proposed to identify him with the enigmatic Horus Sa, 
known from the mention of his Aa-house in inscriptions on stone vessels from the Step 
Pyramid; cf. Lauer, Pyramide V, 7-8, pis. 6-7; Helck, "Die Datierung der Gefassaufschriften 
aus der Djoserpyramide," %AS 106 (1979), 124. 

75 See the excellent photo published in the exhibition catalogue Kemet alle sorgenti del 
tempo, A. M. Donadoni Roveri & F. Tiradritti, eds. (Milan, 1998), 251. 



Fig. II. 2.1. Inscription on stone vessel fragment BM EA 35556 
(drawing after photo: E.-M. Engel) 

Scrutiny of the inscription reveals that the name Ny-netjer is written 
over Weneg. Traces of the plant sign used to write Weneg are dis- 
cernible, as are the enigmatic strokes to the upper left and right of it 
(Fig. II. 2.2) which are also attested in another of his inscriptions. 7b 
Thus Ny-netjer must have been Weneg's successor, and the original 
inscription referred to the palace of Horus Ra c -neb and to nsw bjt 
Weneg. Therefore the Horus name of nsw bjt Weneg should be 

nsw bjt Nub-nefer is attested only on two stone vessels from the 
Step Pyramid (Docs. 23, 24). 7/ Both inscriptions mention him in con- 
nection with a building called Hwt-mn.t- c nh. Another stone-vessel inscrip- 
tion associates this structure with the Gold name Ren.' 8 The Palermo 

Lauer, Pyramide IV. 1, pi. 20: 104; IV.2, 50. 
Lauer, Pyramide IV. 1, pi. VI: 3-4. 
Lauer, Pyramide IV. 1, pi. 19: 98. 



Fig II. 2. 2. Reconstruction of the king's name as originally written on BM EA 35556 

(drawing: E.-M. Engel) 

Stone gives Ren as Ny-netjer's Gold name." Nub-nefer's reign should 
therefore be in proximity to Ny-netjer's. Currently there are two options 
for his dynastic position. Either Nub-nefer was Ra c -neb's nsw bjt 
name"" or he was an ephemeral ruler who occupied the throne briefly 
after Ny-netjer's death. 81 The evidence just presented for identifying 
Weneg as Ra c -neb's nsw bjt name shows that the second alterna- 
tive is correct. Kaiser's suggestion that nsw bjt Sened was the last king 
to reign over UE and LE before Kha-sekhemwy is very plausible. 82 
Circumstantial evidence for this sequence is the survival of the funer- 
ary cult of Sened into Dyn. 4. 83 nsw bjt Sened or nsw bjt Nub-nefer 

79 Recto, IV. 

80 So Gunn, "Inscriptions from the Step Pyramid site III. Fragments of inscribed 
Vessels," ASAE 28 (1928), 156 n. 2; idem, ASAE 44 (1944), 292; cf Beckerath, Handbuch, 48. 

81 Helck (n. 74, Datierung), 131-32; Wilkinson, Egypt, 89. 

82 "Zur Nennung von Sened und Peribsen in Sakkara B 3," GM 122 (1991), 49-55. 
8:i In the tomb of Shery, Saqqara B 3: A. Mariette, Les mastabas de I'ancien empire 

(Paris, 1889), 92-93; A. Moret, "Fragments du mastaba de Shery, pretre des rois 
Peribsen et Send," Monuments Plot 25 (1921/22), 273-98. 


might be identical with Horus Sa, who is known from the mention of 
his ka-house in inscriptions on stone vessels from the Step Pyramid. 84 

It is not clear whether the next two names — Horus Sekhem-ib and 
Seth Per-ibsen — belonged to a single ruler or to two different kings. 
Per-ibsen certainly claimed to rule over all of Egypt, 8 ' 1 but the sources 
do not confirm this. Contemporaneous evidence for Seth Per-ibsen is 
restricted to UE between Elephantine and Beit Khallaf, just north of 
Abydos, 86 except for his funerary cult in association with nsw bjt Sened 
at Saqqara. 87 Sekhem-ib is attested at Abydos and Saqqara. Seal impres- 
sions mentioning Sekhem-ib come from the tomb of Per-ibsen, while 
at Saqqara, stone vessels with Sekhem-ib's name were found in the 
Step Pyramid. But this does not prove that Sekhem-ib exercised influence 
in the Memphite region, since these vessels could have been brought 
to Saqqara from Abydos after Sekhem-ib's death. Theories about the 
relationship between the names Sekhem-ib and Per-ibsen 88 suggest that 
(a) Sekhem-ib and Per-ibsen were names borne simultaneously by a 
single king; 89 (b) Horus Sekhem-ib was the older name of Seth Per- 
ibsen; 90 (c) Horus Sekhem-ib buried Seth Per-ibsen and was thus his 
successor. 91 Down to the present there is no compelling argument favour- 
ing one alternative over the others. 

The last king of Dyn. 2 was Horus-Seth Kha-sekhemwy. His power 
base seems to have been Hierakonpolis where he is attested as victor 
over northern enemies under the name Horus Kha-sekhem. 92 Presumably 

84 Cf. note 74 above and M. J. Raven, "Les fouilles de Leyde dans la tombe de 
Merytneith a Saqqara. Campagnes 2001-2002," BSFE 155 (2002), 31. 

8 '' Kaplony, Inschriften III, fig. 368; cf. Kahl, System, 83-84. 

86 Cf. Kahl, System, 356-58 (Quellen 2887-2914). 

8/ In the tomb of Shery, see note 83, above. Kaiser considers dynastic and political 
reasons that might have led to the institution of Per-ibsen's cult under Kha-sekhemwy. 

88 Cf. Schneider, Lexikon, 405; Wilkinson, Egypt, 90-91. 

89 E.g., Grdseloff, ASAE 44 (1944), 295. 

90 E.g., E. Drioton & J. Vandier, L'Egypte* (Paris, 1962), 164. 

91 E.g., Helck, Thinitenzeit, 103-104. 

92 Cf. the statues Cairo JE 32161 (Quibell, Hierakonpolis I, 11, pis. 40 (upper), 41 
(left); Quibell & Green, Hierakonpolis II (London, 1902), 27-28, 44) and Oxford Ashmolean 
Museum E 517 (Quibell, Hierakonpolis I, 11, pis. 39, 40 (bottom); Quibell & Green, 
Hierakonpolis II, 31, 43-44. See now H. Sourouzian, "L'iconographie du roi dans la 
statuaire des trois premieres dynasties," in: Kunst des Allen Reiches. Symposium im Deutschen 
Archaologischen Institut Kairo am 29. and 30. Oktober 1991 (Mainz: DAIKS 28, 1995), 
141-43, pi. 51. Cf. also the following stone vessels: Cairo CG 14724: Quibell, Hierakonpolis 
I, pi. 38 (upper); P. Kaplony, "Bemerkungen zu einigen Steingefassen mit archaischen 
Konigsnamen," MDAIK 20 (1965), 26 (53), pi. 6; Helck, Thinitenzeit, 72, 106-107; 
J. Baines, Fecundity Figures. Egyptian Personification and the Iconology of a Genre (Warminster, 


later in his reign, Kha-sekhem took the Horus-Seth name Kha-sekhemwy 
to demonstrate that peace and harmony had been restored through his 
actions. 93 Considerable circumstantial evidence exists for Horus Netjery- 
khet as the successor of Kha-sekhemwy: (a) a seal with the names of 
officials used once together with a seal of Kha-sekhemwy and then with 
a seal of Netjery-khet (Doc. 25); (b) Ny-maat-Hep, 94 mother of the royal 
children and king's mother (of Netjery-khet) is attested on seal impres- 
sions in Kha-sekhemwy's tomb at Umm el-Qaab as well as in Tomb 
Kl at Beit Khallaf (Doc. 26a— b); (c) many seal impressions naming 
Netjery-khet which come from Kha-sekhemwy's tomb. 95 

The following Horus names of Dyn. 2 can be associated with nsw 
bjt names or nbw-names: 

Horus Hetep-sekhemwy: nsw bjt Hetep 96 

Horus Ra c -neb: nsw bjt Weneg (see P. 102-103, above) 

Horus Ny-netjer: nsw bjt Ny-netjer, 97 nbw Ren 98 

Horus Sekhem-ib(-Per-en-maat): nsw bjt Sekhem-ib-Per-en-maat" 

Seth Per-ibsen: nsw bjt/nsw bjt Per-ibsen 100 

Horus-Seth Kha-sekhemwy(-Hetep-netjerwy-imef): nsw bjt Kha- 

sekhemwy-Hetep-netjerwy-imef 1( "/Kha-sekhemwy-Nub-khetsen 102 

1985), 245; B.G. Aston, Ancient Egyptian Stone Vessels: Materials and Forms (Heidelberg: 
SAGA 5, 1994), pi. 2a; Philadelphia, Penn. Univ. Mus. E 3958: Quibell, Hierakonpolis 
I, pi. 38 (bottom); Helck, Thinitenzeit 72, 106-107; Baines, Fecundity Figures, 245; Oxford, 
Ashm. Mus. No. not known: Quibell, Hierakonpolis I, pi. 37 (middle, right); Helck, 
Thinitenzeit, 72, 106-107. Presumably also from Hierakonpolis are: Saqqara, no. 662: 
Lauer, Pyramide IV. 1, pi. 3: 18; IV.2, 8 (18); Helck (n. 74, Datierung), 132 n. 7; pri- 
vate collection: Kaplony, MDAIK 20 (1965), 24, 26, fig. 54. 

93 Cf., e.g., Wilkinson, Egypt, 91-92. 

94 See S. Roth, Die Konigsmiitter des Alien Agypten von der Fruhzeit bis zum Ende der 12. 
Dynastie (Wiesbaden: AAT 46, 2001), 59-67. 

to See, for the moment, Dreyer, "Der erste Konig der 3. Dynastie," Stationen, 31—34. 

96 Seal impressions: Kaplony, Inschriften III, figs. 281-82. 

97 Presuming identical Horus and nsw bjt names imply identity; for the latter, 
see Lauer, Pyramide IV. 1, pi. 15: 73. 

98 Palermo Stone, recto, IV. 

99 Again, presuming identical Horus and nsw bjt names imply identity; for the 
latter, see Lauer, Pyramide IV. 1, pi. 18: 87-94. 

100 As the preceeding. For nsw bjt Per-ibsen, cf. Kaplony, Inschriften III, fig. 368; for 
nsw bjt Per-ibsen, cf. Petrie, RT I, pi. 4: 7. 

101 As the preceeding. For nsw bjt Hetep-netjerwy-imef, cf. Kaplony, Inschriften 
III, fig. 214 

102 As the preceeding. For this nsw bjt name, cf. W. M. F. Petrie, Tombs of the 
Courtiers and Oxyrhynkhos (London, 1925J, pi. 8: 18. 


The Lengths of the Reigns of Dyn. 2-kings 

Contemporaneous sources yield comparatively little information about 
the duration of reigns during Dyn. 2. Some stone vessels from the Step 
Pyramid bear inscriptions citing specific events. 10 * One of them men- 
tions the "Fourth Occasion of the Sokar Festival" (probably year 24); 104 
another, the "Seventeenth Occasion (of the cattle count)" (probably 
year 34). 105 These dates have been ascribed to Ny-netjer, since he seems 
to have been the only ruler of the dynasty to have reigned more than 
30 years. 106 For the same reason, inscriptions mentioning a sed festival 
are thought to refer to him. 10 ' The Annals preserve information about 
three kings of Dyn. 2: Ny-netjer, years 6— 21 108 and perhaps x + 9 years 
at the end of his reign; 109 Per-ibsen, 6 + y years from the beginning 
of his reign; 110 Kha-sekhemwy, perhaps years 3— 6 111 and most proba- 
bly years 12-18 towards the end of his reign. 112 

Chronologically Significant Inscriptions, Dyns IS 

Doc. 1 5 clay impressions of a cylinder seal 
Dyn. 1, reign of Den — c Adj-ib 
a-c: Abydos, find nos. Ab.K 300, 301a-b 
d: Abydos 
e: London, UC 188 (provisional no., assigned by Kaplony) 

103 Lauer, Pyramide V, 88-90, (nos. 273-75), figs. 172-74. 

104 Lauer, Pyramide V, 88 (no. 273), fig. 172; Helck (n. 74, Datierung), 128; Wilkinson, 
Egypt, 85-86. 

103 Lauer, Pyramide V, 88-89 (no. 274), fig. 173; Helck (n. 74, Datierung), 128; 
Wilkinson, Egypt, 86. 

106 Cf. Helck (n. 74, Datierung), 128. Wilkinson, Annals, 120, estimates a 39-year 
reign for Ny-netjer. 

107 Cf. Helck (n. 74, Datierung), 130 and Lauer, Pyramide V, 3-7, 59, figs. 6-13: 91. 
But cf. I. Regulski, who assumes a sed festival of Kha-sekhemwy (I. Regulski, "Second 
Dynasty Ink Inscriptions from Saqqara Paralleled in the Abydos Material from The 
Royal Museums of Art and History (RMAH) in Brussels", in: S. Hendrickx, R. F. 
Friedman, K. M. Cialowicz & M. Chlodnicki, eds. Egypt at its Origins (Leuven: OLA 
138, 2004), 960-967. 

108 Palermo Stone, recto, IV. 1-16; cf. Wilkinson, Annals, 119-29, fig. 1. 

109 Cairo fragment 1, recto, IV. 1-9; cf. Wilkinson, Annals, 202-206, fig. 4. 

110 Cairo fragment 1, recto, IV. 10-15; cf. Wilkinson, Annals, 202-203, 207-208, 
fig. 4. 

1 ' ' London fragment, recto, upper part; cf. Wilkinson, Annals, 248-5 1 , fig. 11. 
112 Palermo Stone, recto, V. 1-7; cf. Baud, BSFE 149 (2000), 36-38; Wilkinson, 
Annals, 130-36, fig. 1. 


Table. II. 2.2. The kings from Hetep-sekhemwy to Netjery-khet 

Horus and/or Seth nsw bjt Name Gold Name Sed Festivals 


Hetep-sekhemwy Hetep 

Ra'-neb Weneg 

Ny-netjer Ny-netjer Ren one(?) 

Sa(?) Nub-nefer 


Sekhem-ib Sekhem-ib 

(Per-en-maat) / (Per-en-maat) / 

Per-ibsen (one ruler?) Per-ibsen (one ruler?) 

Kha-sekhem / Kha-sekhemwy 

Kha-sekhemwy (Hetep-netjerwy-imef / 

(Hetep-netjerwy-imef) Nub-khetsen) 

Netjery-khet Netjery-khet Nub(?) 

Prov.: a-d, Umm el-Qaab, Tomb T; e, not known but prob- 
ably also Tomb T. 

Comment: See Pp. 96-97, above 

Bibliography: Kaplony, Inschriften III, fig. 809; Dreyer, MDAIK 43 (1987), 

33-43, pis. 3-5; Kaiser, MDAIK 43 (1987), 115-19; Dreyer, MDAIK 

49 (1993), 61 

Doc. 2 several clay impressions of a cylinder seal 
Dyn. 2, reign of Hetep-sekhemwy 
Abydos, find. nos. Ab.K 1486-1500 
Prov.: Umm el-Qaab, Tomb Q_ 

Comment: See Pp. 96-97, above 

Bibliography: Dreyer, MDAIK 52 (1996), 72-73, fig. 26, pi. 14b-c 

Doc. 3 2 fragments of a serpentine bowl 
Dyn. 1, reign of Djer — "Serpent" 

Berlin, Egyptian Mus. 15453 and Abydos, find no. Ab.K 
Prov.: Umm el-Qaab Tomb Z and Tomb T, respectively 

Comment: These fragments bear incised Horus names attesting the seg- 
ment Djer — "Serpent" of the sequence established by Doc. 1. 
Bibliography: Petrie, RT II, 25, pi. 7: 1; V. Miiller et al, "Umm el- 


Qaab. Nachuntersuchungen im friihzeitlichen Konigsfriedhof. 11./ 12. 
Vorbericht", MDAIK 56 (2000), 109-10, fig. 22a, pi. 10a. 

Doc. 4 schist bowl 

Dyn.l, reign of Den, with inscriptions added under c Adj- 

ib — Qa- C a 

Cairo, Egyptian Museum JE 88345 

Prov.: Saqqara, Step Pyramid, Gallery B 

Comment: The bowl bears four incised nsw bjt/nsw bjt names: 
Khasty, <Mer>-pi-bia, Iry-netjer, and Qa- C a. 

Bibliography: Lauer, Pyramide III, pi. 19: 2; IV. 1, 10, pis. I: 7, 4: 21; 
IV.2, 10-12. 

Doc. 5 6 fragments of a schist plate 

Dyn. 1, reign of Den, with inscriptions added under c Adj- 

ib — Q_a- C a 

Prov.: Saqqara, Step Pyramid, Gallery H 

Comment: The inscriptions record the nsw bjt/nsw bjt names Khasty, 
[Mer-pi-]bia, Iry-netjer and <Qa-> c a. 

Bibliography: Firth and Quibell, The Step Pyramid II (Cairo, 1935), pi. 
105: 3; Lauer, Pyramide IV. 1, 9-10, pi. 4: 20; IV.2, 9-12; Helck, 
Thinitenzeit, 113, 187. 

Doc. 6 2 fragments of a schist bowl 

Dyn. 1, reign of Den, with inscriptions added under c Adj- 

ib — Qa- C a 

Cairo, Egyptian Museum JE 55254-55255 

Prov.: Saqqara, Step Pyramid, outside Galleries H and B 

Comment: The adjoining fragments bear four incised nsw bjt/nsw bjt names, Khasty, Mer-pi-bia, Iry-netjer and Qa- C a. 
Bibliography: Gunn, ASAE 28 (1928), 156, 158, pi. I: 2; Firth and 
Quibell, Step Pyramid II, pi. 88: 1; Lauer, Pyramide IV. 1, 10, pi. 4: 19; 
IV.2, 9-12; Helck, Thinitenzeit, 113, 187. 

Doc. 7 calcite bowl 

Dyn. 1, reign of Den, with inscriptions added under c Adj- 
ib — Qa- C a 

Michailidis collection 
Prov.: not known 


Comment: The bowl bears the nsw bjt/nsw bjt names Khasty, Mer- 
pi-bia, Iry-netjer and Qa- C a. 

Bibliography: Kaplony, Beschriftete Kkinfunde in der Sammlung Georges Michai- 
lidis. Ergebnisse einer Bestandsaufnahme im Sommer 1968 (Istanbul, 1973), 6 
(25), pi. 7: 25. 

Doc. 8 "pyroxen-syenit" bowl 

Dyn. 1, reign of Den, with inscriptions added under c Adj- 

ib — Qa- C a 

Swiss private collection 

Prov.: not known 

Comment: The bowl shows four incised nsw bjt/nsw bjt names: 

Khasty, Mer-pi-bia, Iry-netjer and Qa- C a. 

Bibliography: Kaplony, Steingefdsse, 20-24 (9), pis. 2, 17, 18 

Doc. 9 "black stone" (diorite?) bowl 

Dyn. 1, reign of Den, with added inscriptions of c Adj-ib and 


New York, MMA 58.125.2 

Prov.: not known (ex. Michailidis collection) 

Comment: The bowl shows three inscised nsw bjt/nsw bjt names: 

Khasty, Mer-pi-bia and Iry-netjer. 

Bibliography: Kaplony, MDAIK 20 (1965), 13 (22), pis. 3: 22, 4: 22. 

Doc. 10 fragment of a rock crystal bowl 

Dyn. 1, reign of Den, with added inscriptions of c Adj-ib and 


London, BM EA 49278 

Prov.: Umm el-Qaab, Tomb T 

Comment: Mer-pi-bia, originally incised near the name Khasty (untouched), 
was erased and replaced by Iry-netjer, documenting the sequence Khasty, 
Mer-pi-bia, Iry-netjer. 

Bibliography: E. Naville, The Cemeteries at Abydos I (London, 1914), 35, 
pis. 8 (lower right), 14: 1; Spencer, Objects, 42 (271), pis. 23: 271, 26: 271. 

Doc. 1 1 fragment of a stone vessel 

Dyn. 1, reign of Den, with added inscriptions of c Adj-ib and 


DYNASTIES O— 2 1 1 1 

Paris, Louvre E. 11035 

Prov.: Umm el-Qaab, Tomb T 

Comment: As Doc. 10. 

Bibliography: E. Amelineau, Les nouvelles fouilles d'Abydos 1895—1896, 
(Paris 1899), pi. 42; K. Sethe, "Die altesten geschichtlichen Denkmaler 
der Agypter", %AS 35 (1897), 3; Lauer, Pyramide IV.2, 10. 

Doc. 12 fragment of a quartz-crystal bowl 

Dyn. 1, reign of Den, with added inscription of c Adj-ib 
Philadelphia, Univ. Mus. E 06847 
Prov.: Umm el-Qaab, Tomb U 

Comment: The names Khasty and Mer[-pi]-bia are incised on the 


Bibliography: Petrie, RT I, 19, 38-39, pi. 5: 9; II, pi. 47: 31. 

Doc. 13 fragment of a red limestone bowl 

Dyn. 1, reign of Den, with added inscription of c Adj-ib 
Cairo, Egyptian Museum JE 34378 
Prov.: Umm el-Qaab, Tomb T 

Comment: The vessel may have originally shown more names than 

Khasty and Mer-pi-bia. 

Bibliography: Petrie, RT I, 19, 38-39, pi. 5: 12. 

Doc. 14 three adjoining fragments of a calcite vessel 

Dyn. 1, reign of Den with added inscription of Adj-ib 

London, BM EA 32659 

Prov: Umm el-Qaab, Tomb X or Z 113 

Comment: c Adj-ib's Horus name is written in front of the nsw bjt name 


Bibliography: Petrie, RT I, 19, 38, pi. 5: 11; Spencer, Objects, 41 (268), 

pi. 25: 268. 

Doc. 15 three adjoining fragments of a schist bowl 

Dyn. 1, prior to the reign of Semer-khet, with inscriptions 
added naming Semer-khet and Qa- C a 

Cf. Spencer, Objects, 41 (268). 


Cairo, Egyptian Museum JE 55268 (one fragment only) 
Prov.: Saqqara, Step Pyramid, outside Galleries H and B 

Comment: The field in the serekh has been cut down, i.e. the name 
Semer-khet has been substituted for another. 

Bibliography: Gunn, ASAE 28 (1928), 158, pi. 1: 5; Lauer, Pyramide 
IV. 1, pis. IV: 3, 8: 38; IV.2, 22, (38); Helck, Thinitenzeit, 101-102. 

Doc. 16 part of a schist bowl 

Dyn. 1, reign of c Adj-ib, with added inscription of Qa- C a 
Prov.: Saqqara, Step Pyramid, Great South Court 

Comment: The two nsw bjt names incised on the vessel are Mer- 
pi-bia and Qa- C a; Iry-netjer was not mentioned. 
Bibliography: Lauer, Pyramide IV. 1, pi. 8: 36; IV.2, 21 (36). 

Doc. 17 schist bowl 

Dyn. 1 , reign of Semer-khet with added inscription of Qa- C a 
Cairo, Egyptian Museum JE 88344 
Prov.: Saqqara, Step Pyramid, Gallery B 

Comment: The treatment of the serekhs shows very well that different 
hands inscribed the two Horus names Semer-khet and Qa- C a. 
Bibliography: Lauer, Pyramide IV.l, pi. 8: 39; IV.2, 22; Kaplony, Inschriften 
I, 593. 

Doc. 18 granite statue of a kneeling man 
Dyn. 2 or 3 

Cairo, Egyptian Museum CG 1 
Prov.: Mit Rahineh 

Comment: The sign preceding the names Hetep-sekhemwy, Ra c -neb 
and Ny-netjer incised in that order on the right shoulder blade of the 
figure has been interpreted to read divine ancestor (Helck), falcon (Fischer), 
or phoenix (Moret). 

Bibliography: PM III (2nd ed.), 864; A. Moret, "L'influence du decor 
solaire sur la pyramide," in: Melanges Maspero I (Cairo, 1961), 624, fig. 1; 
Fischer, Artibus Asiae 24 (1961), 45-46; E. L. B. Terrace & H. G. Fischer, 
Treasures of Egyptian Art from the Cairo Museum (London, 1970), 25-28; 
Helck, Thinitenzeit, 240. 

DYNASTIES O— 2 1 1 3 

Doc. 19 diorite bowl 

Dyn. 2, reign of Hetep-sekhemwy, with added inscription of 

Ra c -neb 

Cairo, Egyptian Museum JE 65413 

Prov.: Saqqara, Step Pyramid, Gallery H 

Comment: The Horus names of Hetep-sekhemwy and Ra c -neb, incised 
by different hands, face the goddess Bastet. 

Bibliography: Lauer, Pyramide IV. 1, pis. II: 8, 11: 58; IV.2, 31-32; 
Fischer, Artibus Asiae 24 (1961), 46-47. 

Doc. 20 flint bowl 

Dyn. 2, reign of Hetep-sekhemwy (or earlier) with inscrip- 
tions naming Hetep-sekhemwy and Ra c -neb 
Cairo, Egyptian Museum JE 41981 
Prov.: Giza, Valley Temple of Mycerinus 

Comment: Both Horus names, incised by different hands, face the god- 
dess Bastet. Hetep-sekhemwy is incised over an erasure, while Ra c -neb, 
partially erased, is written behind Hetep-sekhemwy. Reisner argued that 
Ra c -neb was succeeded by Hetep-sekhemwy on the basis of the partial 
erasure of Ra c -neb's name, but Docs. 21 and 22 show that Ra c -neb 
was Hetep-sekhemwy's successor. 

Bibliography: G. A. Reisner, Mycerinus. The Temples of the Third Pyramid 
at Giza (Cambridge/Mass., 1931), 102 (1), 179 (1), pi. 70c; Fischer, 
Artibus Asiae 24 (1961), 46-47; Helck, Thinitenzeit, 72, 103. 

Doc. 21 footed schist bowl 

Dyn. 2, reign of Hetep-sekhemwy or Ra c -neb, with added 

inscription of Ny-netjer. 

Prov.: Saqqara, Step Pyramid, Gallery B 

Comment: The bowl provides evidence for the reign of Ny-netjer being 
later than that of Hetep-sekhemwy. A reference to the "A:a-house of 
Horus Hetep-sekhemwy" follows the name Ny-netjer. As Lacau and 
Lauer noted, Ny-netjer could have been substituted for the name of 
another king (Hetep-sekhemwy or Ra c -neb.) 

Bibliography: Lauer, Pyramide IV. 1, pi. 15: 74; IV.2, 36 (74); Helck, 
Thinitenzeit 195-196. 


Doc. 22 fragment from the rim of a bowl of volcanic ash 

Dyn. 2, reign of Ra c -neb, with added inscription of Ny- 


London, BM EA 35556 

Prov.: Umm el-Qaab, Tomb P 

Comment: See Pp. 102—103, above 

Bibliography: Petrie, RT II, 26, 51, pi. 8: 12; R. Weill, La He et la Ilk 

dynasties (Paris, 1908), 150; Spencer, Objects, 42 (275), pi. 26: 275; Helck, 
Thinitenzeit, 103, 195-96 (n. 130); Spencer, in: Kemet (see n. 75, above), 

Doc. 23 fragment of a schist bowl 

Dyn. 2, reign of Ny-netjer or slightly later 

Cairo, Egyptian Museum JE 55268 

Prov. Saqqara, Step Pyramid, outside Galleries H and B 

Comment: nsw bjt Nub-nefer is mentioned in connection with the build- 
ing Hwt-mn.t- c nh, also attested from the reign of Ny-netjer. 
Bibliography: Lauer, Pyramide IV. 1, 6, pi. VI: 3; IV.2, 48-49. 

Doc. 24 fragment from the rim of a schist bowl 

Dyn. 2, reign of Ny-netjer or slightly later 

Cairo, Egyptian Museum JE 55294 

Prov.: Saqqara, Step Pyramid, outside Galleries H and B 

Comment: nsw bjt Nub-nefer is mentioned in connection with the build- 
ing hwt-mn.t-nh; cf. Doc. 23. 
Bibliography: Lauer, Pyramide IV. 1, 6, pi. VI: 3; IV.2, 48-49. 

Doc. 25 clay seal impressions 

Dyn. 2, reign of Kha-sekhemwy 

Oxford, Ashm. Mus. A 1909.1 1 18A-D, F-O; London, UC 

112-114 (nos. assigned by Kaplony) 

Prov.: Abydos, Shunet ez-Zebib 

Comment: The seal (Kaplony, Inschriften III, fig. 346) with names of 
officials, which made these impressions was used in tandem with a seal 
of Kha-sekhemwy as well as with another naming Netjery-khet; cf. 
P. 106, above. 

Bibliography: Kaplony, Inschriften I, 164—66; II, 869 n. 1011; ///, figs. 
269, 346 and 801. 

DYNASTIES O— 2 1 1 5 

Doc. 26a clay seal impressions 

Dyn. 2 or 3, reign of Kha-sekhemwy or Netjery-khet 
Cairo, Egyptian Museum CG 11106-112, 11143, 11145; 
Chateaudun 38(2) (no. assigned by Kaplony); Paris, Louvre 
no no.; Toronto, ROM B 2324; London, UC 92(2), 95, 96 
(nos. assigned by Kaplony); Abydos (cf. Dreyer, infra) 
Prov.: Umm el-Qaab; Tomb V 

Comment: These impressions from the tomb of Kha-sekhemwy men- 
tion Queen Ny-maat-Hep as mother of the royal children. 
Bibliography: Kaplony, Inschriften I, 161; II, 855 (983), 1137 (325); ///, 
fig. 325; Dreyer, Stationen (see n. 95 above), 33. 

Doc. 26b clay seal impressions 

Dyn. 3, reign of Netjery-khet 

London, UC 149-52 (nos. assigned by Kaplony) 

Prov.: Beit Khallaf, Tomb Kl 

Comment: These impressions from a tomb dating to the reign of Netjery- 
khet name Queen Ny-maat-Hep as king's mother. Therefore one can 
conclude for a sequence Kha-sekhemwy — Netjery-khet. 
Bibliography: Kaplony, Inschriften /, 167; //, 873 (1070), 1137 (326); ///, 
fig. 326. 


Stephan J. Seidlmayer 

To reconstruct the chronology of Dyn. 3, 1 the number, names and 
sequence of its kings must first be established. Very little information 
can be gleaned from the Manethonian tradition since the surviving 
epitomes are unfortunately marred by erratic repetitions and inflated 
reign lengths. 2 But the kinglists of the NK 3 do provide crucial evidence, 
while pWestcar preserves the names and sequence of two Dyn. 3 rulers. 4 
The data from these sources, summarized in the table below, must be 
correlated with each other, and with OK evidence: 

Westcar TC 

Saqqara list 

Abydos list 



Nb-hi (19 yrs.) 



(28 yrs.) 


Dsr Dsrjt (19 yrs.) 




(29 yrs.) 


Dsrtj (6 yrs.) 



Tyreis (var. 
Tyris; 7 yrs.) 

six more kings 

Hw-dfi (6 yrs.) 



(17 yrs.) 





(16 yrs.) 

(19 yrs.) 

(continued on next page) 

1 Recent discussions include: Baud, "Menes"; idem, Djeser et la Hie dynastie (Paris, 
2002); Beckerath, Chronologie, 160-163; G. Dreyer, "Der erste Konig der 3. Dynastie", 
in: Stationen, 31-34; S. Seidlmayer, "Die staatliche Anlage der 3. Dynastie", in M. Bietak, 
ed., House and palace in Ancient Egypt (Vienna, 1996), 198-200 note 14. 

2 Waddell, Manetho, 40—45. For the relationship between Manetho and the kinglists 
of the NK, see Helck, Manetho, 19-24 and Beckerath, Chronologie, 160. 

'■'' TC, col III, lines 4-8; Saqqara list, nos. 12-15; Abydos list, nos. 15-19. 
4 A. M. Blackman, The Story of King Kheops and the Magicians, Repr., W. V. Davies, 
ed., (Reading, 1988), col. 1.14, col. 1.19 and passim. 

DYNASTY 3 117 

Table (cont.) 

Westcar TC Saqqara List Abydos List Manetho Manetho 

(Africanus) (Eusebius) 

Hwnj (24 yrs.) Hwnj Aches 

(42 yrs.) 


(30 yrs.) 


(26 yrs.) 

total: 214 yrs. total: 198 yrs. 
(aram: 197 yrs.) 

Associating names from the lists with contemporaneous monuments — 
above all, with royal mortuary complexes — provides additional infor- 
mation about the number and sequence of kings. In OK sources, the 
names of only two Dyn. 3 kings were written in cartouches: Nebka 
and Huni. Normally, the kings of the dynasty were identified in con- 
temporaneous sources by their Horus names, not by their personal 
names. Therefore, the problem of correlating Dyn. 3 attestations with 
the names of the later lists arises. 

Five Horus names are known with certainty from Dyn. 3 contexts: 
Netjery-khet, Sekhem-khet, Kha'ba, Zanakht and Qahedjet. Two other 
names were considered by Swelim. 5 Ink inscriptions on some stone ves- 
sels from the galleries below the Step Pyramid, 6 i.e., in a Dyn. 2 con- 
text,' attest the ^a-house of za. Equating za with Horus Zanakht and 
identifying him as Djoser's predecessor and the owner of the original 
mastaba below the Step Pyramid 8 seems unfounded. 9 ^a and Zanakht 

5 Swelim, Problems, 181-183. 

6 Lacau & Lauer, Pyramide V, 7; B. Gunn, "Inscriptions from the Step Pyramid Site", 
ASAE 28 (1928), 168. 

' W. Helck, "Die Datierung der Gefassaufschriften aus der Djoserpyramide", JfAg 
106 (1979), 120-132. 

8 So J.-P. Lauer, Observations sur les pyramides (Cairo: BdE 30, 1960), 82; Kaplony, 
Inschriften I, 409; J.-P. Lauer & H. Altenmiiller, in: C. Vandersleyen, Das Alte Agppten 
(Berlin: Propylaen Kunstgeschichte 15, 1975), 113. 

9 Doubts were expressed by: Beckerath, Handbuch, 177; Kaplony, Steingefdsse, 7 n. 8, 
and Wildung, Rolle, 55; Helck rejected these ideas, initially in "Datierung" (n. 7), 130, 
and subsequently in Thinitenzeit, 108. 


are clearly different; the common phonetic element za is written with 
a different sign in each name. Since za never occurs alone or in a 
serekh, it is doubtful that it represents a king's name. 10 Reports on an 
alleged Horus B>, mentioned by Swelim, cannot be verified; even if he 
could be shown to exist, there are no grounds for assigning him to 
Dyn. 3. 

Reviewing the available documentation, a number of problems are 
evident which are taken up in turn below. Both TC and the Abydos 
List name Nebka as the first ruler of the dynasty. But in pWestcar, a 
king Nebka is a successor of Djoser, while a basically similar name 
occurs in the Saqqara List as the penultimate entry and as the last in 
the Abydos List. It was always suspected that there was only a single 
king Nebka whose name was duplicated (in the Abydos List) or shifted 
from its correct position. Since all attestations for Nebka from the OK 
can be shown to refer to a king who reigned near the end of Dyn. 3, 
the existence of a like -named king at the beginning of the dynasty was 
always doubtful. Recent excavations at Abydos revealed unequivocal 
evidence that Horus Netjery-khet buried Kha c -sekhemwy, the last king 
of Dyn. 2," there, making it certain that no reign (and especially, a 
chronologically significant one as shown in TC) could have intervened 
between them. 

Djoser, the second name in the list and, in fact, the first king of 
Dyn. 3, is well attested. Later inscriptions confirm his identity with 
Horus Netjery-khet, the owner of the Step Pyramid complex at Saqqara. 12 
His successor Djoserti/Djoser-teti is certainly Horus Sekhem-khet, the 
owner of the second, smaller step pyramid complex at Saqqara. Evidence 
for the equation includes the morphological similarity of the two archi- 
tectural complexes, their geographic proximity and the fact that the 
Nebti-Name of Sekhem-khet is attested as Djosert(i)- c ankh on an ivory 
plaque from his pyramid. 15 

Data for the three remaining kings are less easily brought into line. 
The so-called Layer Pyramid of Zawyet el-Aryan is the only other pyra- 

10 See J.-P. Lauer, "A propos de l'invention de la pierre de taille", in: Fs G. Mokhtar 
(Cairo: BdE 97.2, 1985), 62-63. 

11 Dreyer, in: Stationen, 31-33. 

12 The relevant sources are the Famine stela at Sehel, temp. Ptolemy V (P. Barguet, 
La stele de la famine a Sehel, Cairo: BdE 24, 1953), a statue dedicated by Ahmose (Berlin 
14765; Wildung, Rolle, 79-83) and numerous graffiti left by NK visitors to the Step 
Pyramid complex (Wildung, Rolle, 65-72). 

13 M. Baud, Djeser (n. 1), 67—68. Doubts expressed by Helck, "Das Kleidertafelchen 
aus der Pyramide des Shm-hf, WZKM 54 (1957), 72-76, do not seem justified. 

DYNASTY 3 119 

mid complex datable to Dyn. 3. 14 The "great excavation" at the same 
site is not contemporaneous with the Layer Pyramid, but belongs instead 
to later Dyn. 4. 15 The brickwork-complex of el-Der, discussed by Swelim, lb 
is probably not a pyramid at all (although it is difficult to judge this 
monument on the basis of available documentation). 1 ' If it were a pyra- 
mid, its comparatively small size would associate it with the series of 
small step pyramids in UE. Finally, evidence currently available sup- 
ports the assignment of the initial building phase of the pyramid at 
Maidum to Snofru (and Dyn. 4), not to Huni. 18 

Architecturally, the Layer Pyramid of Zawyet el-Aryan is very close 
to the step pyramids at Saqqara — for example, the extensive subter- 
ranean magazines are a common feature — and quite different from the 
earlier phase of the Maidum pyramid. Accordingly, the structure is 
most likely assignable to the successor of Horus Sekhem-khet/Djoserti; 
in any case, it should not belong to Huni. The archaeological record 
preserves no evidence of mortuary complexes for two later Dyn. 3 
kings, a fact which may not be due simply to accident of preservation. 
Their absence may reflect instead a temporary restructuring of arrange- 
ments for the royal mortuary cult related to the appearance at the end 
of the dynasty of the series of small step pyramids in UE associated 
with the names of kings Nebka, Huni and Snofru. 19 This line of rea- 
soning supports an early date for the Layer Pyramid. 

The name of the Layer Pyramid's owner, Djoser's second successor, 
is listed as missing (Sds, Hw-dff) in TC. Several stone vessels from 

14 PM III, 313. 

15 Convincingly established by J.-P. Lauer, "Sur 1'age et l'attribution de l'excavation 
monumentale a Zaouiet el-Aryan", RdE 14 (1962), 21-36. Thus the much-discussed 
royal name which appears in the masons' graffiti from this monument (e.g. J. Cerny, 
"Name of the King of the Unfinished Pyramid at Zawjet el-Aryan", MDAIK 16 (1958), 
25-29, and A. M. Dodson, "King [Bjk-b]", Z$S 108 (1981), 171) is irrelevant for the 
history and chronology of Dyn. 3. 

16 Swelim, Problems, 337. 

17 Cf. the archaeological situation as presented in the original publication, M.R. 
Macramallah, "Une forteresse du moyen empire (?) a Abou Rawach", ASAE 32 (1932), 
161-173. Macramallah's opinion on the dating of the site is very tentative. 

18 See R. Stadelmann, "Snofru und die Pyramiden von Maidum und Dahschur", 
MDAIK 36 (1980), 443-446. 

19 W. Kaiser & G. Dreyer, "Zu den kleinen Stufenpyramiden Ober- und Mittel- 
agyptens", MDAIK 36 (1980), 43-59; Seidlmayer, in "Anlage" (n. 1), 205-209; idem, 
"Town and State in the Early OK, a view from Elephantine", in Aspects of Early Egypt, 
A.J. Spencer, ed., (London, 1996), 108-127. Horus Zanakht/Nebka is represented by 
a seal impression in the context of the royal complex attached to the pyramid at 
Elephantine and the name Huni occurs in an inscription on a block from the pyra- 
mid; for Snofru as the owner of the pyramid of Seila see J. Leclant, Orientalia 57 (1988), 
336 Fig. 40, and Orientalia 58 (1989), 368. 


mastaba Z500 in the cemetery of Zawyet el-Aryan 20 bear the Horus 
name Kha'ba who might well be the owner of the pyramid. But, the- 
oretically, Kha'ba could also be a successor of the Layer Pyramid's 
owner. As the next ruler, the NK lists name Nebka-Re c or Nefer-ka- 
Re c , both corrupted from Nebka, the name attested in pWestcar and 
twice in the OK. The earlier OK attestation occurs in inscriptions from 
the mastaba of Akhet'a who held a priestly title in the king's cult. 21 
Helck identified Akhet'a with a person whose titles (but neither the title 
relating to the cult of Nebka nor the name) appear among the inscrip- 
tions on stone vessels from the Step Pyramid, to conclude that both 
the person and King Nebka date to early Dyn. 3. 22 But this equation 
is neither obvious nor convincing. Even if Helck were correct, Akhet'a 
could have nevertheless outlived Djoser and his immediate successors 
to serve in the cult of a later king Nebka. The style of the reliefs from 
Akhet'a's tomb and the overall similarity of its plan to the tomb chapel 
of Metjen clearly point to a late Dyn. 3/early Dyn. 4 date for the 
owner. The fact that the king's name in Akhet'a's title is written in a 
cartouche clearly precludes the possibility that the tomb could have 
been decorated under Djoser. 

The second OK attestation of King Nebka occurs in the name of a 
funerary domain associated with the cult of Neusserre c . 2,i The context 
links Nebka to late Dyn. 3/early Dyn. 4 when the system of funerary 
domains is first attested, under Huni and Snofru. 24 Thus the pre-NK 
sources substantiate the presence of a king Nebka towards the end of 
Dyn. 3, a conclusion quite in keeping with the fact that the existence 
of a like-named ruler at the very beginning of Dyn. 3 can be excluded. 

Significantly, Nebka can be identified with Horus Zanakht on the 
basis of a seal impression from mastaba K2 at Beit Khallaf where the 
names occur in parallel. 25 The king is attested as Horus Zanakht at a 

20 For the inscriptions see D. Dunham, Ramjet el-Aryan. The Cemeteries of the Layer 
Pyramid, (Boston, 1978), 29-34, pi. XIIA. G. A. Reisner, Mycerinus, The Temples of the 
Third Pyramid at Giza, (Cambridge Mass., 1931), 103, classifies Z500 as a stairway 
mastaba, making Swelim's idea (Problems, 78-89) that it could be the mortuary tem- 
ple of the Layer Pyramid unlikely. 

21 Ziegler, Catalogue, 96-103. 

22 W. Helck, "Datierung" (n. 7), 129; idem, Thinitenzeit, 107. 

23 Borchardt, Ne-user-Re, 79 Fig. 54. 

24 H. K. Jacquet-Gordon, Les noms des domaines funeraires sous I'Ancien Empire Egyptien 
(Cairo: BdE 34, 1962), 7-10. 

25 Seidlmayer, in "Town" (n. 19), 121, and PL 23. 

DYNASTY 3 121 

number of other sites, in the rock inscriptions of Wadi Maghara, and 
by seal impressions from the mortuary temple of the Step Pyramid and 
from Elephantine. 26 One sealing from Elephantine derives from layers 
of refuse in the royal complex attached to the small step pyramid at 
the site establishing a connection between Nebka and the series of small 
step pyramids in UE. Unfortunately, the stratigraphic position of the 
sealing is too equivocal to allow the relationship to be defined more 
precisely. 27 

As the last ruler of Dyn. 3, the TC and the Saqqara List cite Huni, 
a corrupted form of nswt Hwj.' m Huni was apparently the first king reg- 
ularly identified in contemporaneous sources by his personal name, pref- 
aced nswt (by contrast to the later usage of nswt-bjt), written in a 
cartouche. To date, no document gives Huni's Horus name in associ- 
ation with his personal name. There is, however, one monument which 
needs to be discussed in this context: a relief slab of Horus Qahedjet 
depicting him embraced by Horus of Heliopolis. 29 Allegedly from 
Dahshur, it should derive from a Dyn. 3 royal mortuary complex at 
the site. In fact, two alabaster sarcophagi from a shaft tomb in the 
vicinity of the mortuary precinct of Senwosret III provide evidence for 
the presence of a Dyn. 3 royal complex at Dahshur. 30 Accordingly, 
Qahedjet is an ideal candidate for Huni's Horus name. It should not 
be overlooked, however, that the correlation is not absolutely certain. 
If Kha'ba is not the name of the owner of the Layer Pyramid at Zawyet 
el-Aryan, but rather of one of his successors, then Kha'ba could qualify 
as Huni's Horus name. (Because of its early architectural type, it is 

26 Discussed by Seidlmayer, in W. Kaiser et al., "Stadt und Tempel von Elephantine, 
9./ 10. Grabungsbericht", MDAIK 38 (1982), 304 with n. 82. 

27 Ibidem, 304-305, idem, in "Anlage" (n. 1), 198-200. 

28 This reading was established by L. Borchardt, "Konig Huni?", %AS 46 (1909), 
12-13, and corroborated by H. Schafer, "Konig Huni", £45 52 (1914), 98-100. The 
discussion triggered by H. Goedicke's article "The Pharaoh Ny-swth", %AS 81 (1956), 
18-24, and continued by E. Meltzer, "A reconsideration of [njswt Hwj]" , JEA 57 (1971), 
202-203, W. Barta, "Zum altagyptischen Namen des Konigs Aches", MDAIK 29 (1973), 
1-4, and W. Helck, "Der Name des letzten Konigs der 3. Dyn. und die Stadt Ehnas", 
SAK 4 (1976), 125-134, remains unconvincing. Why should not the first king who reg- 
ularly had his personal name inscribed in a cartouche include the title nswt inside the 
ring, in contrast to what became convention only subsequendy? Actually this would 
parallel the usage attested much more frequendy with the epithet z3-R c which was 
sometimes included within the cartouche. 

29 J. Vandier, "Une stele egyptienne portant un nouveau nom royal de la troisieme 
dynastie", CRAIBL (1968), 16-22; Ziegler, Catalogue, 54-57. 

30 PM III, 885. I am grateful to Dieter Arnold for pointing this out to me. 



unlikely that the pyramid is Huni's tomb). If so, the name Qahedjet 
could be the Horus name of Djoser's second successor whose personal 
name is lost from the NK lists. Since Qahedjet probably owned a mor- 
tuary complex at Dahshur, the Layer Pyramid would necessarily belong 
to Horus Zanakht/Nebka. This reconstruction is not in itself impossi- 
ble, but it seems to suit the available evidence less well. The results of 
these deliberations may be summarized as follows: 

Name in cartouche 

Horus name 

Mortuary complex 



great step pyramid, Saqqara 



lesser step pyramid, Saqqara 


H c -bl 

layer pyramid, Zawyet el- Aryan 




nswt Hwj 


mortuary cult place of unknown type 
at Dahshur 

Data currendy available are sufficient neither for determining the length 
of reigns for each king nor the length of the dynasty as a whole. Nor 
does contemporaneous evidence exist for estimating a minimum length 
of reign for any king. Only TC provides more or less useful data. 
However, King Nebka was displaced in this document to the begin- 
ning of the dynasty; perhaps the 1 9-year reign accorded him does not 
represent original information but simply duplicates the figure given for 
Djoser. The lengths of reigns in the Manethonian tradition are arbi- 
trarily inflated; it would be guesswork to attempt to reduce them just 
as arbitrarily. 

The Old Kingdom Annals unfortunately can provide little assistance. 51 
The Palermo fragment preserves the end of Kha c -sekhemwy's reign and 
the beginning of Djoser's in line 5. In addition, Kaiser noticed that, 
taking all the available criteria into account, the distance between Cairo 
1 and the Palermo fragment can be reconstructed so that the distance 
between the two changes of reign preserved on both fragments matches 
the combined reign lengths of Djoser and Djoserti as recorded in the 
TC:' 2 While the line separating the two reigns is not actually visible on 

31 See Beckerath, Chronologic, 174-179; Baud, "Menes", 135-138; idem, Djeser (n. 1), 
50-52; Wilkinson, Annals. 

:i2 W. Kaiser, ZAS 86 (1991), 44 with note 8. 

DYNASTY 3 123 

the stone, 33 such a reconstruction would still seem possible, in view of 
the worn state of the surface at this place. Therefore, the evidence of the 
Annals at least does not necessarily contradict the testimony of the TC. 
Any reconstruction of the lengths of the reigns for later Dyn. 3 rulers 
depends on estimating the distance between Cairo fragment 1 and the 
left edge of the original stone slab. Beckerath's recent calculation allows 
only 24 years for all three remaining kings, in clear contradiction of 
TC which gives 24 years to Huni alone plus six more for "Hu-djefa". 
In addition, an unknown number of years has to be taken into account 
for Nebka. In view of the extremely fragmentary state of the evidence, 
it would be hasardous to put excessive trust into any attempt to recon- 
struct the original slab. Given the current state of knowledge, it is wiser 
to regard the total of 50 years for the entire dynasty, which emerges 
from Beckerath's reconstruction, only as a minum figure. However, 
even if the 19 years for Nebka of the TC be considered reliable, the 
total length of the dynasty would not exceed 75 years. 

Wilkinson, Annals, 53. 


Miroslav Verner 

The crucial problem for reconstructing the relative chronology of Dyns. 
4 and 5 is our ignorance of how the system of dating according to the 
"year/occasion of the (cattle) count" (rnpt zp) was employed. Ever since 
Gardiner's paper, 1 specialists have acknowledged that this census formed 
the basis for counting regnal years during the OK. The existence and 
use of the term rnpt m-ht zp ("year after the count") was taken as clear- 
cut evidence for a biennial census. However, the preponderance of rnpt 
Zp over rnpt m-ht zp years among the burgeoning number of dates recov- 
ered in recent years from excavations, taken in conjunction with an 
entry on the Palermo Stone attesting the cattle count in two succes- 
sive years of Snofru, indicates that the situation is much more com- 
plex. Nowadays, some Egyptologists maintain that the census was biennial 
during Snofru's reign with the sole exception of the seventh and eighth 
counts which were conducted in successive years. 2 Others are of the 
opinion that a biennial system was not employed under Snofru, 5 while 
yet others equivocate. 4 For subsequent reigns, opinions fluctuate from 
the presumption of a biennial system 5 to the assumption that on 
certain occasions an "odd" count could have been ignored. 6 Finally, 
there is also the theory that annual cattle counts became more and 
more frequent during the OK until they became the rule by the end 
of Dyn. 6. 7 

Obviously the existence of the census per se is not at issue, but rather 
its regularity during the OK. However, a statistical review of documented 

1 Gardiner, "Years", 11-28. 

2 E.g. R. Stadelmann, "Beitrage zur Geschichte des Alten Reiches", MDAIK 43 
(1987), 229-239; Baud, "Menes", 121. 

3 E.g. R. Krauss, "Pyramid", 47—50. 

4 E.g. Spalinger, "Texts", 281. 

5 E.g. Baud, "Menes", 122-123, 128-129. 

6 Spalinger, "Texts", 318. 

' Beckerath, Chronologie, 147. 


dates, even if the list is incomplete and the attribution of some con- 
tested, allows some conclusions. For example, from the beginning of 
Dyn. 4 to the end of Dyn. 5, the years of the count and those fol- 
lowing the count occurred in succession, and rnpt zp years were con- 
sistently more frequently documented than rnpt m-ht zp years. Moreover, 
no clear-cut tendency can be observed towards a marked decrease in 
the number of years following the count throughout the period. 8 On 
the contrary, the evidence for rnpt zp and rnpt m-ht zp from Snofru's 
reign at the beginning of Dyn. 4 to Djedkare c 's at the end of Dyn. 5 
supports the opposite conclusion. 

Could a cattle count take place in the year of a king's accession to 
the throne? Until quite recently, the opinion prevailed that the first cattie 
count of any given king's reign occurred during the first full year after 
his accession. But an entry among the annals inscribed on the stone 
recently discovered at South Saqqara casts doubt on this assumption, 
since the text explicitely mentions a (cattle) count of the year of the 
"Unification of the Two Lands" at the beginning of the reign of Merenre'. 9 

The potential usefulness of the series, though incomplete, of rnpt zp 
and rnpt m-ht zp dates for Dyns. 4 and 5 is demonstrated by analysis 
of the data for both Snofru and Djedkare c . Down to the present, the 
highest recorded number of rnpt zp years for Snofru is 24; only half of 
them (viz., rnpt zp 2, 7, 8, 12, 13, 14, 15, 16, 18, 23 and 24) are doc- 
umented among the preserved dates. Can we assume that about the 
same proportion — i.e., ca. half of the evidence for the intervening rnpt 
m-ht zp years — remains to date unattested? If so, the number (three) of 
such currently documented dates — rnpt m-ht zp 10, 13 and 18 — should 
be doubled. An estimate for the length of Snofru's reign based on these 
data would be 24 + 6 = 30 years. Using other arguments, Krauss 10 
and, independently, Barta 11 arrived at nearly the same length for the 
king's reign. 

Djedkare's highest documented census is the 21st (or possibly the 
22nd). Thirteen other "occasions" are known: 1, 3, 4(?), 5, 6, 8, 9, 10, 
11, 14, 15, 16, 17, and 18. In this series, seven (and possibly eight) 

8 Contra Beckerath, Chronologic 

9 Baud & Dobrev, "Annales", 47. 

10 R. Krauss, "The Length of Sneferu's reign and how long it took to build the 
Red Pyramid", JEA 82 (1996), 48. 

11 W. Barta, "Die Chronologie der 1. bis 5. Dynastie nach den Angaben des rekon- 
struierten Annalensteins", %iS 108 (1981), 21. 


mpt zp or about one-third of the total is still unattested. There are 
seven mpt m-ht zp dates preserved: 1, 3, 4, 7(?), 10, 14, and 17. Adding a 
third of this amount, or about two years, to the number of mpt zp (21 
or 22?) and mpt m-ht zp (7) years actually attested yields 30 (or possibly 
31) years for the reign. It must be emphasized, however, that for the 
present these calculations, since speculative, must be treated with reserve. 

No matter how cautiously conclusions be drawn, the available data 
indicate that the dating system was irregular during Dyns. 4 and 5, 
and not principally biennial with few exceptions. In practice, annual 
cattle counts apparently prevailed. The theory that intervening years 
were omitted from the record under certain circumstances 12 is contra- 
dicted by the so-called masons' inscriptions which consistently refer only 
to mpt zp years. These short texts associated with the construction pro- 
jects of the state are the most frequently preserved dated documents 
from Dyn. 4 and 5. Why should these inscriptions regularly omit every 
second year from the administrative record? 

If an irregular dating system pertained during Dyns. 4 and 5, what 
economic and/or administrative necessity determined its irregularity? It 
is possible, for instance, that during the formative period of the bureau- 
cracy the frequency of the census was linked to the financing of large 
projects — buildings, reclamation of land from the marshes, etc. Were 
consecutive census years occasioned by funding shortfalls? Were factors 
influencing the decision to organize the census annually or biennally 
always the same or did they differ? 

A special problem is the discrepancy between the available contem- 
poraneous evidence and the reign lengths recorded of Dyns. 4 and 5 
rulers in the TC. By contrast to the Manethonian tradition, the TC 
was long considered by specialists to be the standard against which 
contemporaneous data should be measured. Reign lengths given in the 
papyrus were sometimes used as evidence for annual cattle counts and 
at other times, for a biennial census. The compiler was presumed to 
have omitted one or more signs from some figures and to have mis- 
takenly duplicated entries. Obviously, comparison of data from the very 
damaged papyrus with contemporaneous evidence can hardly be expected 
to provide a definitive version of OK chronology. The names of only 
three Dyn. 4 and 5 kings survive in the papyrus out of a presumed 
17; three more partially-preserved names can be reconstructed. The 

12 Posener-Krieger, "Graffiti"; see also Spalinger, "Texts", 318. 



remaining eleven are lost. Furthermore, the sequences at the end of 
Dyn. 4 and in mid-Dyn. 5 are by no means certain. Nor does evi- 
dence from contemporaneous documents always inspire confidence. For 
example, there is only a single case where the precise date of a king's 
death and the accession of his successor are known, viz. for Sahure c 
followed by Neferirkare c . Regardless, the exact length of Sahure"s tenure 
still cannot be established, since we do not know how regularly the 
census was taken during his reign. 

Disappointing as this may be, analysis of contemporaneous dates, both 
rnpt zp and rnpt m-ht zp, offers some stimulating insights. Assuming that 
the census was irregular throughout Dyns. 4 and 5, the minimum length 
of a given king's reign should equal the total of the highest attested cen- 
sus year (rnpt zp) and the number of attested intervening years (rnpt 
m-ht zp)- A comparison of the results of such calculations with the data 
recorded in the TC is represented in the table that follows: 

Contemporaneous Evidence 


Dyn. 4 


27 + x years 

Cheops (Khufu 

13 + x 


11 (10?) + x 

Khephren (Ra'kha'ef) 

15 + x 


not attested 

Mycerinus (Menkaure 1 ) 

14 (?) + x 


2 + x 


not attested 

Dyn. 5 


4 + x 


8 + x 


5 + x 


not attested 


1 + X 


8(?) + x 


not attested 


28 (29 ?) + x 


9 + x 

24 years 














+ x 

As can be seen, the estimate for the length of Snofru's reign exceeds 
the figure provided by the papyrus; the same is probably true for the 
reigns of Ra'djedef and Djedkare c . On the other hand, the entry in 


the TC for Shepseskare c , if correctly associated with that king, does not 
reflect the contemporaneous record, since virtually no clear-cut evi- 
dence for this mysterious ruler has yet surfaced. 13 Such discrepancies 
call into question the credibility of the papyrus for Dyns. 4 and 5. 14 

List of Contemporaneously Documented Dates, Dyns. 4—5 


TC 111.9: 24 years 

Manetho: Soris — 29 years 

Beckerath, Chronologie, 159: 35 years (?) 

rnpt zp 
2 (n) tnwt; Cairo frg no. 4 15 

7 (n) tnwt; Palermo Stone, recto 6 16 
7, ibd III; Maidum pyramid 17 

8 (18 ?), ibd III smw, sw 2; Maidum pyramid 1 " 
8 (n) tnwt; Palermo Stone, recto 7 19 

12, ibd IV smw, sw 1 (?); Maidum pyramid 20 

13, . . . prt (?), sw 10; Maidum pyramid 21 
13 (16 ?), ibd I smw, sw . . .; Maidum pyramid 22 
13, ... smw, sw . . .; Maidum pyramid 23 
13 (16 ?), ... smw, sw . . .; Maidum pyramid 24 
13; Maidum pyramid 25 

13 M. Verner, "Who was Shepseskare, and when did he reign?", in: M. Barta, 
J. Krejci, eds., Abusir and Saqqara in the year 2000 (Prague, 2000), 581-602. 

14 Verner, "Remarks". 

15 H. Gauthier, "Quatre nouveaux fragments de la pierre de Palerme", in: G. Maspero, 
Le Musee egyptien III (Cairo, 1915), 50-52. 

16 Schafer, "Annalen", 30. 

17 Posener-Krieger, "Graffiti", 20, pi. 8, A.20. 

18 Posener-Krieger, "Graffiti", pi. 8, A.27. 

19 Schafer, "Annalen", 31. 

20 Posener-Krieger, "Graffiti", pi. 8, A.33. 

21 Posener-Krieger, "Graffiti", pi. 7, A. 1. 

22 Posener-Krieger, "Graffiti", pi. 7, A. 11. 

23 Posener-Krieger, "Graffiti", pi. 7, A.5. 

24 Posener-Krieger, "Graffiti", pi. 7, A.9. 

25 Posener-Krieger, "Graffiti", pi. 7, A.2. 


13; Maidum pyramid 26 

14 (17 ?), ibd II smw, sw . . .; Maidum pyramid 2 ' 

14 (17 ?), . . . prt; Maidum pyramid 28 

14 (17 ?), ibd I + x; Maidum pyramid 29 

15, ibd II prt, sw 14; Dahshur, Red Pyramid 30 

15 (?), ibd III prt; Maidum pyramid 31 

15, ibd III smw, sw 10 + x; Maidum pyramid 32 
15 (?), ibd IV smw (?), sw 10; Maidum pyramid 33 

15 (?), ... smw (?), sw . . .; Maidum pyramid 34 
15; Dahshur, Red Pyramid 3 ' 

16, ibd I iht, sw 13; Dahshur, quarry mark 36 
16, ibd III iht; Dahshur, Red Pyramid 3 ' 

16, ibd IV iht, sw 14; Maidum pyramid 38 

16, ibd II (?) smw, sw 12; Maidum pyramid 39 

16 (?), ibd . . .prt, sw 2; Dahshur, Red Pyramid 40 
16 (?), ibd I prt; Maidum pyramid 41 

16 (?), ibd III prt; Maidum pyramid 42 
16; Maidum pyramid 43 

17, ibd II prt, sw 10 + x; Maidum pyramid 44 
17, ibd I prt, sw 20; Maidum pyramid*' 

■ 8, 


■ 8, 


■ 8, 


■ 8, 


26 Posener-Krieger, "Graffiti" 

27 Posener-Krieger, "Graffiti" 

28 Posener-Krieger, "Graffiti" 

29 Posener-Krieger, "Graffiti" 

30 R. Stadelmann (n. 2), 234-235, fig. 2. 

31 Posener-Krieger, "Graffiti", pi. 7, A.4. 

32 W. M. F. Petrie, E. J. H. Mackay & G. A. Wainwright, Meydum and Memphis III 
(London, 1910), 9, pi. V, 6. 

33 Posener-Krieger, "Graffiti", pi. 7, A.6. 

34 Posener-Krieger, "Graffiti", pi. 7, A.9. 

35 Stadelmann (n. 2), 233-5, fig. 1. 

36 LD II, I g. 

37 Stadelman (n. 2), 234-235, fig. 2. 

38 A. Rowe, 77k Museum Journal 22 (1931), 26, pi. 38, fig. 2. 

39 Posener-Krieger, "Graffiti", pi. 7, A.3. 

40 H. Sourouzian, MDAIK 38 (1982), 389-390, fig. 5. 

41 Rowe (n. 38), 26. 

42 Posener-Krieger, "Graffiti", pi. 8, A.22. 

43 Posener-Krieger, "Graffiti", pi. 7, A.7. 

44 Petrie et al. (n. 32), 9, pi. 5, 2 left. 

45 Posener-Krieger, "Graffiti", pi. 7, A. 13. 



17, 'bd I prt, sw 22; Maidum pyramid 46 
17, Ibd III prt, sw c rq; Maidum pyramid 47 
17, ibd III prt, sw . . .; Maidum pyramid 48 
17, ?bd III 3ht, . . .; Maidum pyramid 49 
17, . . .prt; Maidum pyramid'" 
17, ('bd) I + x prt; Maidum pyramid' 1 
17, (ibd) I + x; Maidum pyramid 52 
17, ibd. . .; Maidum pyramid 53 
17; Maidum pyramid 54 

'bd I prt, sw 21; Maidum pyramid' 5 

'bd II smw; Maidum pyramid' 6 

ibd II iht, . . . (?); Dahshur, Red Pyramid 5 ' 

'bd . . .prt, . . .; Dahshur, Red Pyramid 58 



Damaged evidence of rnpt zp 

10 + x, 'bd IV smw; Maidum pyramid 59 

10 + x; Maidum pyramid 60 

(1)6 (?), 'bd I + x, sw 12; Maidum pyramid 61 

(1)6 (?), ibd I + x, smw (?), sw 2; Maidum pyramid 62 

(1)7 (?), IV smw, sw 21; Maidum pyramid 63 

. . ., ibd IV . . ., sw . . .; Maidum pyramid 64 

46 Posener-Krieger, 


, pi. 7, 

A. 14. 

47 Petrie et al. (n. 32), 9, pi. 


48 Petrie et al. (n. 32), 9, pi. 

5, 3. 

49 Posener-Krieger, 


pi. 7, 

A. 12. 

50 Posener-Krieger, 


, pi. 7, 

A. 16. 

51 Posener-Krieger, 


, pi. 7, 

A. 17. 

52 Posener-Krieger, 


, pi. 7, 

A. 18. 

53 Posener-Krieger, 


pi. 8, 


54 Posener-Krieger, 


pi. 7, 

A. 19. 

55 Posener-Krieger, 


, pi. 8, 


56 Posener-Krieger, 


, pi. 9, 


57 Reconstruction of 

a mason's 


in LD 

fig. 3. 

58 Stadelmann (n. 2), 239-240, fig. 


59 Posener-Krieger, 


, pl- 8, 

A. 34. 

1,0 Posener-Krieger, 


, pl. 8, 


61 Posener-Krieger, 


pl. 7, 


62 Posener-Krieger, 


pl. 7, 

A. 10. 

63 Posener-Krieger, 


pl. 7, 

A. 15. 

1,4 Posener-Krieger, 


, pl. 9, 


2), 234-236, 


• . . ., ibd III smw, sw . . .; Maidum pyramid 65 

• . . ., ibd III smw, . . . 10 + x; Maidum pyramid 66 

rnpt (m)- xt zp 67 

• 10, ibd 1 + x; Maidum pyramid 6 " 

• 13, . . .; Maidum pyramid 69 

• 18, ibd IV smw, sw (?) 5; Maidum pyramid' 

damaged evidence of rnpt (m-)ht zp 
'...', Maidum pyramid' 1 

attested rnpt zp: 2; 7; 8; 12; 13; 14; 15; 16; 17; 18; 23; 24 
attested rnpt (m-)ht zp'- 10; 13; 18 
rnpt zp: rnpt (m-)ht zp — 12: 3 

rCIII. 10 (?): 23 years 
Manetho: Siiphis (I) — 63 years 
Beckerath, Chronologie, 159: 23 years 

rnpt zp 

• 4 (?), ibd.. .; G 2130, Khentika 72 

• 5, . . . smw (?), sw 5; G 1203 73 

• 8, ibd I prt, . . . (?); a loose (?) block found at the upper end of the 
causeway, near the entrance to the king's mortuary temple 74 

• 8, ibd III smw, sw 20; G 4000, Hemiunu 75 

• 10, ibd IV prt, sw 23 (or 24); G 4000, Hemiunu 76 

65 Posener-Krieger, "Graffiti", pi. 9, A.37. 

66 Posener-Krieger, "Graffiti", pi. 9, A.41. 

67 For the transcription of m-ht see Edel, Grammatik, 180. 

68 Posener-Krieger, "Graffiti"," pi. 8, A.30. 

69 Posener-Krieger, "Graffiti", pi. 8, A.32. 

70 Posener-Krieger, "Graffiti", pi. 8, A.28. 

71 Posener-Krieger, "Graffiti", pi. 9, A.38. 

72 Attributed to Khufu by Smith, "Evidence", 118 fig. 6; 127 no. 4; so also Y. Harpur, 
Decoration in Egyptian Tombs of the Old Kingdom (London, 1987), 269. 

73 Attributed to Khufu by Smith, "Evidence", 118, fig. 6; 127 no. 2, and by G. A. 
Reisner, A History of the Giza Necropolis I (Cambridge, Mass., 1942), 76 n. 2 and 391/19/. 

74 Attributed to Khufu by Smith, "Evidence", 119 fig. 7; 126f. no. 1; originally, 
A. Rowe read this date "year 13", see Reisner (n. 73), 71. 

75 Attributed to Khufu by Junker, Giza I, 159, fig. 24/10/, 161. 

76 Junker, Giza I, 161, no. 12. 


• 10, ibd I smw, sw 10 + x; G 4000, Hemiunu 7 ' 

• 10, ibd II smw, sw 10 + x; G 4000, Hemiunu 78 

• 12, ibd II smw; G 2120, Seshatsekhentiu 79 

• 12, ibd II...; G 7130-40, Khufukhaf I 80 

rnpt (m-)ht zp 

• 13 K1 

attested rnpt zp: 4, 5, 8, 10, 12 

rnpt (m-)ht zp'- 13 

rnpt zp: rnpt (m-)ht zp — 5: 1 


rCIII. 11 (?): 8 years 

Beckerath, Chronologie, 159: 9 years 

rnpt zp 

• 1, ibd III prt, . . .; pyramid of Ra'djedef 82 

• 11 (or 10 ?), ibd I pr(t), sw 24 (?); boat pit, south side of Kh epos's 
pyramid 83 

attested rnpt zp: 1, 11 (10 ?) 
rnpt (m-)ht zp: not attested 
rnpt zp: rnpt (m-)ht zp: 2: (?) 

77 Junker, Giza I, 158, 160; 159 fig. 24/1/. 

78 Junker, Giza I, 159 fig. 24/2/; 160. 

79 Attributed to Khufu by Smith, "Evidence", 1 18, fig. 6; 127 no. 3, and by Spalinger, 
"Texts", 285; according to N. Strudwick, The Administration of Egypt in the Old Kingdom 
(London, 1985), 117 no. 6, the reign of Khephren is also possible. 

80 Attributed to Khufu by Smith, "Evidence", 119, fig. 7; 127 no. 8, and by W. K. 
Simpson, The Mastabas of Kawab, Khafkhufu I and II (Boston, 1978), 9. This dating also 
accords with Stadelmann's theory that Khufukhaf I assumed the name Khephren after 
succeeding Ra'djedef, see SAK 11 (1985), 165-172. 

81 K. P. Kuhlmann, in: Tides of the Desert — Gezeiten der Wiiste. Contributions to the 
Archaeology and Environmental History of Africa in Honmour of Rudolph Kuper (Koln: Africa 
Praehistorica 14, 2002), 125-138. 

82 M. Vallogia, in: Etudes sur lAncien Empire et la necropole de Saqqara dedies a Jean- 
Philippe Lauer (Montpellier, 1997), 419. 

8:i According to I. E. S. Edwards, in: The Unbroken Reed: Studies in the Culture and 
Heritage of Ancient Egypt in honour of A. F. Shore ( London, 1994), 101, 105 n. 20, Posener- 
Krieger read "year 10" (i.e. "year of the 10th cattle count"); see also R. Krauss, 
Orientalia 66 (1997), 3 n. 16. Both readings are possible; however, the reading of the 
left column with the date is not doubt-free, and the right column of the graffito is 
almost illegible, see A. M. Abubakr & A. Y. Mustafa, The Funerary Boat of Khufu, in: 
BABA 12 (1971), 11, fig. 6 bottom left. 


rCIII. 12: 20 + x years 
Manetho: Suphis (II) 66 years 
Beckerath, Chronologie, 159: 26 years 

rnpt zp 

• 1, ibd IV iht, sw 5; ostracon from Helwan tomb 299 H 2 84 

• 5, ibd III prt, sw 22; ostracon from Helwan tomb 335 H 2 85 

• 7, ibd IV prt, sw 10; G 7530-40, Meresankh IIP 6 

• 7, ibd IV prt, sw 20; G 7530-40, Meresankh IIP 7 

• 10, ibd III smw, sw 24; ostracon Leiden J 429 88 

• 10 (?), ibd III smw, ...; G 7350, Hetepheres II (?) K9 

• 12, ibd II smw, sw 10; G 7650, Akhtihotep and his wife Meretites 90 

84 Z. Saad, Royal Excavations at Saqqara and Helwan (1941-1945), Suppl. ASAE no. 3 
(Cairo, 1947), 106, pi. 42 a left; as Spalinger notes, "Texts", 287, Khephren's car- 
touche in the inscription clinches its assignment to his reign. 

85 Saad (n. 84), 107, pi. 43 at right; probably temp, of Khephren. 

86 Attributed to Khephren by D. Dunham & W. K. Simpson, The Mastaba of Mersyankh 
///(Boston, 1974), 3 fig. 1; see also Smith, "Evidence", 127 no. 9, 119 fig. 7. 

87 Attributed to Khephren by Dunham & Simpson (n. 86), 3 fig. 1; see also Smith, 
"Evidence", 127 no. 9, 119 fig. 7. Anthropological examination of Meresankh Ill's 
bones put her age at death at about fifty, see V. G. Callender, Egypt in the Old Kingdom 
(Melbourne, 1998), 172-173. 

88 According to H. Goedicke, JEA 54 (1968), 24, 28-29, pi. 5 no. 4, the ostracon 
is of the same date as those of Helwan; thus rnpt zp ^0 probably refers to Khephren. 
Cf. also, idem, Old Hieratic Palaeography (Baltimore, 1988), pi. 16. 

89 According to Reisner (n. 73), 73 n. 2; see also Smith, "Evidence", 119 fig. 7, 127 
no. 9. The date was inscribed on the back of a block from G 7350, purportedly built 
for Hetepheres II. However, Smith identified the figures in the relief on the front as 
Hetepheres II (?) and Meresankh III (?), and he dated it to the time of Shepseskaf 
(HESPOK, 1946, 164-165, 302, pi. 45 a). The attribution of G 7350 to Hetepheres II 
is not based on textual evidence. P. Janosi, %AS 123 (1996), 56-57, has questioned 
Smith's conclusions. He suggests that the relief might have originally shown an anony- 
mous prince followed by his mother and his wife. No doubt the attribution of the date 
is fraught with difficulties, regardless of the fact that a high date and Shepseskaf 's 
reign are mutually exclusive, which leaves either Khephren or Menkaure'. Since 
Meresankh III was probably buried in the tomb intended originally for her mother at 
the beginning of Menkaure"s reign (see the dates rnpt zp 1 and rnpt (m-)ht zp 1 sub 
Menkaure c ), it would be surprising to find her represented with her mother in a tomb 
built as late as Menkaure"s reign. When considering the data from G 7350 and G 
7530-40 and with due circumspection in view of the complex stratigraphy and unclear 
chronology of cemetery G 7000, one is inclined to assign the date to Khephren. 

90 Attributed to Khephren by Smith, "Evidence", 119 fig. 7, 127-128 no. 11 b; see 
also Reisner (n. 73), 73 n. 1. For doubts about the attribution to Khephren, and a 
still higher date of rnpt zp 1 3 (see below) from the tomb of Akhtihotep, see P. Janosi, 
Giza in der 4. Dynastie. (Wien 2005), 71-73, 443. 


• 12 Tnwt.. .; LG 87, Nikaure 91 

• 13, ibd TV. . .; G 7650, Akhtihotep and his wife Meretites 92 

rnpt (m-)ht zp 

• 4 Tnwt, >bd II smw, sw 3; ostracon from Helwan tomb 305 H 2 93 

• 4 Tnwt, >bd II smw, sw 4; ostracon from Helwan tomb 305 H 2 94 

• 5, ibd II smw, sw 8; ostracon from Helwan tomb 322 H 2 9a 

• 5, ?bd III prt, sw 22; ostracon from Helwan tomb 335 H 2 % 

attested rnpt zp: 1, 5, 7, 10, 12, 13 
attested rnpt (m-)ht zp'- 4, 5 
rnpt zp: rnpt (m-)ht zp — 6: 2 


rCIII. 13 (?):...?... years 

Manetho: 22 years (Bicheris, preceded by Ratoises, was inserted by 

Manetho between Mycerinus and Shepseskaf.) 

Beckerath, Chronologie, 159: 7 years 

rCIII. 14 (?): 18 (28 ?) years 
Manetho: Mencheres — 63 years 
Beckerath, Chronologie, 159: 28 years. 

91 Urk. I, 16.14; see also H. Goedicke, Die privaten Rechtsinschriften aus dem Alien Reich 
(Wien, 1970), 21-22. There is no unanimity on the dating of the will of Khephren's 
son Nikaura. For instance, Strudwick (n. 79; 107) concluded that Nikaura should have 
been born in Khephren's reign, "and thus would be no older than twenty-two at the 
end of his father's reign". Consequently, rnpt zp 12 should then apply to Khephren's 
successor Menkaure'. With reference to art historical criteria and the replacement of 
Tnwt by ipt in the date (the former being supposed by Goedicke, Rechtsinschriften, 22, 
to have disappeared by the beginning of Dyn. 5), Spalinger ("Texts", 294) opts for 
Menkaure', too. But Baud, Mines, 128, argues that Khephren's name occurs in Nikaura's 
tomb with such an insistence that the date should refer to this king. Janosi (n. 90), 
too, does not exclude the attribution of the date to Khephren, provided that Nikaura 
was born before his father ascended the throne. Taking all the arguments in account, 
including the possibility that Khephren might have become king later in his life, one 
is inclined to assign this date to him. 

92 Attributed to Khephren by Smith, "Evidence", 119 fig. 7, 128 no. 11. 

93 Saad (n. 84), 106-107, pi. 42 b right. For the translation and interpretation, see 
H. G. Fischer, Orientalia 29 (1960), 187-190; Spalinger, "Texts", 287. 

94 Saad (n. 84), 106-107, pi. 42 b left. For translation and the reference of the ostra- 
con, see Fischer (n. 93), 187-90; see also Spalinger, "Texts", 287. 

95 Saad (n. 84), 107, pi. 43 a right; see Spalinger, "Texts", 288. 

96 Saad (n. 84), 106-107, pi. 43 b right; see Spalinger, "Texts". 


rnpt zp 

• 2, ibd II prt, sw 22; G VI S 97 

• 2, ibd IV smw, sw 22; G 7530-40, Meresankh IIP 8 

• 11, ibd..., sw 10 + x; G VI S" 

The following dates from the Gebelein papyri can probably be attrib- 
uted to Mycerinus: 10 ° 

rnpt (m-)ht zp 

• 2, ibd . . . iht, sw 20; Gebelein, frag. A 101 

• 3, ibd III prt, sw 26; Gebelein, frag. B 102 

• 11, . . .; Gebelein, rouleau IV 103 

attested rnpt zp'- 2, 1 1 

attested rnpt (m-)ht zp: 2(?), 3(?), 11(?) 

rnpt zp: rnpt (m-)ht zp — 2: 3(?) 

rCIII. 15 (?): 4 years 
Manetho: Sebercheres — 7 years 
Beckerath, Chronologie, 159: 5 years 

97 Junker, Giza X, 75, fig. 35.9, 78, no. 10. With regard to the occurrence of this 
date on some blocks at the site, Junker attributed the date to Menkaure' because of 
the presence of his crew names. 

98 Dunham & Simpson, (n. 86), 3, fig. 1 e; see also Smith, "Evidence", 116 fig. 4, 
126 no. 2. The date is inscribed to the north of the subsidiary niche, eastern facade, 
of the mastaba (from the date on the south side, only . . . prt, sw 1 7 survived). Reisner 
attributed the date to Khephren, see Smith, "Evidence"; Spalinger, "Texts", 286, accepts 
this dating, though with some hesitation. However, Reisner's dating can be seriously 
questioned. If the tomb was built around the 7th census of Khephren, as indicated by 
two masons' inscriptions (Simpson & Dunham (n. 86), 3, Fig. 1 b, c), the lower date 
found on the mastaba's facade can hardly be earlier. But to which event did the date 
refer? The attribution of the date to Menkaure' seems to be, therefore, more proba- 
ble. For the complex, and the problem of its history, see Janosi (n. 90), 500 and idem, 
ZM 123 (1996), 46-62. 

99 Junker, Giza X, 75 fig. 35.10, 77 no. 9; because Menkaure"s crew names were 
found on some blocks at the site, Junker attributed the date to his reign. 

100 P. Posener-Krieger, RdE 27 (1975), 216 and idem in: Fs Elmar Edel: 12. Marz 1979 
(Bamberg, 1979), 318-331. 

101 Posener-Krieger (n. 100; 1979), 318-331. 

102 Posener-Krieger, RdE 27 (1975), 215-216. 

103 Posener-Krieger, RdE 27 (1975), 215-216. 


mpt zrrS tiwy 

• Palermo Stone 104 

• ibd II smw, sw 7#; 1(b 

• ibd II smw, sw 10; G 5552 106 

• ibd III smw, sw...; G 7450 107 

• ibd IV smw, sw 4; G 7450 108 

mpt zp 

• 1, ibd 1 smw, sw 21 (G 753040, Meresankh III) 109 

mpt (m-)ht zp 

• tpy, ibd II prt, sw 28 110 

• 1 (n) ipt (!) ih c wt nbt, edict of Shepseskaf for the pyramid of Mycerinus 1 

attested mpt zp'- 1 
attested mpt (m-)ht zp'- 1 
mpt zp: mpt (m-)ht zp — 1: 1 

TC III, 16 (?): 2 years 
Manetho: Thamphthis — 9 years 
Beckerath, Chronologie, 159: 2 years 


rCIII. 17: 7 years 

Manetho: Usercheres — 7 years 

Beckerath, Chronologie, 159: 8 years 

104 Schafer, "Annalen", 32-33. 

105 Attributed to Shepseskaf by Helck, in: Fs Goedicke, 107. 

106 Attributed to Shepseskaf by Helck (n. 105), 107-108. 

107 Attributed to Shepseskaf by Helck (n. 105), 107-108. 

108 The date to the right of the entrance to Meresankh Ill's chapel refers to the 
preparation of the queen's burial. Reisner attributed it to Shepseskaf, see Smith, "Evi- 
dence", 126, fig. 4, 118; Janosi (n. 90), 501 concurs. But Dunham & Simpson (n. 86), 
8, pi. 2 a, fig. 2, and also Spalinger, "Texts", 288-289, assign it to Menkaure'. 

109 The date, inscribed on the left side of the entrance to Meresankh Ill's chapel 
and referring to the completion of the queen's burial, was attributed by Reisner to 
Shepseskaf, see Smith, "Evidence", 126, fig. 4 on p. 118. Janosi (n. 90), 501, attrib- 
utes the date to Shepseskaf, too. On the other hand, Dunham & Simpson (n. 86), 8, 
pi. 2 a, and also Spalinger, "Texts", 289, attribute the date to Menkaure'. 

110 G. A. Reisner, Mycerinus: The temples of the third Dynasty at Giza (Cambridge, Mass., 
1931), 278 no. 1; see also H. Goedicke, Konigliche Dokumente axis dem Altai Reich (Wiesbaden: 
AA 4, 1967), 16-17, fig. 1 and P. Janosi, GM 141 (1994), 49-54. 

111 Schafer, "Annalen", 34. 


rnpt zp 

• 3; Palermo Stone, verso 2 112 

• 3, ibd III prt, sw . . .; sun temple of Userkaf 113 

rnpt (m)-xt zp 

• 1 (n) tnwt; Cairo frag. no. 1 recto 2 114 

attested rnpt zp'- 3 
attested rnpt (m)-xt zp'- 1 
rnpt zp: rnpt (m)-xt zp, 1: 1 


TC III 18 (?): 12 years 
Manetho: Sephres — 13 years 
Beckerath, Chronologie, 155: 13 years 

rnpt zp 

• 1; Cairo Frg. no. 1 verso 2 115 

• 2, ibd I smw, sw 20; mason's inscription, mortuary temple of Sahure c116 

• 4, ibd IV >ht, sw 12; masons' inscription, mortuary temple of Sahure tU7 

• 5, >bd I >ht; sun temple of Userkaf, tablet A 118 

• 5, >bd III prt; sun temple of Userkaf, tablet B 119 

• 5, >bd III smw; sun temple of Userkaf, tablet C 120 

rnpt (m)-xt zp 

• 2 (Palermo Stone, verso 3) 121 

• 5, ibd II prt; sun temple of Userkaf, tablet D 122 

• 6; Palermo Stone, verso 4 123 

112 Schafer, "Annalen", 34. 

113 Haeny, in: BABA 8 (1969), 41-42 no. 6. 

114 Gauthier (n. 15), 45-46, pi. 26. 

115 Gauthier (n. 15), 47. 

116 L. Borchardt, Das Grabdenkmal des Konigs Sahure I, (Leipzig, 1910), 88, M 26. 

117 Borchardt (n. 116), 89, M 29. 

118 Probably to be attributed to Sahure', see Verner, "Remarks", 386-390. 

119 Probably to be attributed to Sahure', see Verner, "Remarks", 386-390. 

120 Probably to be attributed to Sahure', "Remarks", 386-390. 

121 Schafer, "Annalen", 36-37. 

122 Probably to be attributed to Sahure', see Verner, "Remarks", 386-390. 

123 Schafer, "Annalen", 38-39. To date, this is the highest contemporaneous date 
attested for Sahure"s reign. A. Roccati, La litterature historique sous lAncien Empire (Paris, 
1982), 48, for instance, read the worn signs as 7 and, provided the census was bien- 
nial, evidence for year 15 of Sahure"s reign. However, as pointed out by Wilkinson 
(Annals, 168), 6 is a more likely reading than 7; in case of a biennial census, it would 
refer to year 13. 


attested rnpt zp'- 1, 2, 4, 5 
attested rnpt (m)-xt zp'- 2, 5, 6 
rnpt zp: rnpt (m)-xt zp, 4: 3 


TC III. 19 (?): length of reign lost 
Manetho: Nefercheres — 20 years 
Beckerath, Chronologie, 155: 20 years 

rnpt zm? Sivy; Palermo Stone verso 4 124 

rnpt zp 

• 5; Palermo Stone verso 5 125 

• 5, >bd TV lift, sw 4; mason's inscription, pyramid of Khentkaus II 126 

• 5, >bd TV; mason's inscription, pyramid of Neferirkare cl2/ 

attested rnpt zp'- 5 

mpt m-ht zp'- not attested 

rnpt zp'- rnpt m-ht zp — 1: 

TC 111.20 (?): 7 years 
Manetho: Sisires — 7 years 
Beckerath, Chronologie, 155: 7 years 


TC 111.21 (?): 1 year 

Manetho: Cheres — 20 years 

Beckerath, Chronologie, 155: 11 years 

rnpt zp 

• tpy, >bd IV >ht sw 4 + x; pyramid of Ra'neferef 128 

attested rnpt zp'- 1 

rnpt (m-)ht zp'- not attested 

mpt zp: rnpt (m-)ht zp — 1: 

124 Schafer, "Annalen", 39. 
12> Schafer, "Annalen", 40. 

126 Probably to be attributed to Neferirkare', see Verner, Z^ s 10 7 (1980), 159, 
fig. 3; idem The Pyramid Complex of Khentkaus (Prague, 1995), 43-45. 

127 L. Borchardt, Das Grabdenkmal des Konigs Nefer-ir-ke'-re' (Leipzig, 1909), 46 and n. 6. 

128 Corrected copy: Verner, ^AS 126 (1999), 76, fig. 6. 



TC 111.22 (?): 11 (+ x years ?) 

Manetho: Rathures — 44 years 

Beckerath, Chronologie, 155: 31 years; cf. ibidem 208, where the figure 

30 — or 20 ? + 1 or 5 ? — is given 

rnpt zm3 (t3wy) 

• unpublished potsherd no. 763/1/84— x, mortuary temple of Ra'neferef 

rnpt zp 

• 1, ibd I iht, sw 10 + x; unpublished potsherd no. 531/1/82, mortu- 
ary temple of Ra'neferef 

• 1, ibd tpy iht, . . .; unpublished potsherd no. 763/1/84— e, mortuary 
temple of Ra'neferef 

• 2, ibd III smw, sw 10; a potsherd found by Borchardt 129 in (or to the 
west of ?) the mastaba of Djadjamankh in Abusir 

• 5 (?), >bd III prt, (sw) wpw; masons' inscription, mastaba of Ptahshepses 130 

• 7, ibd III >ht, sw 1 (or 7 ?); jar for beef fat no. 531/1/82, mortuary 
temple of Ra'neferef 

rnpt (m-)ht zp 

• 2, ibd 3 iht, sw 24; masonry block, found in the southern "Eckbau" 
of the mortuary temple of Neuserre' 131 

attested rnpt zp: 1, 2, 5(?), 7 
attested rnpt (m-)ht zp'- 2 
rnpt zp: rnpt (m-)ht zp — 4(?): 1 


TC III. 23: 8 years 

Manetho: Mencheres — 9 years 

Beckerath, Chronologie, 155: 9 years 


TC III. 24: 28 years 

Manetho: Tancheres — 44 years 

Beckerath, Chronologie, 155: 38 years 

129 L. Borchardt, Das Grabdenkmal des Konigs Me-user-r' (Leipzig, 1907), 139. 

130 To be attributed to Neuserre': M. Verner, Baugrqffiti der Ptahschepses Mastaba (Prague, 
1992), 110, graffito no. 194. 

131 Borchardt (n. 129), 145. 


mpt zm> Swy 

• ibd III prt, sw 29; tomb of Wepemneferet 


mpt zp 

• 1, ibd II iht, . . .; unp. pap., (mortuary temple) archive of Ra'neferef, 
pi. 51 A 

• mpt tpy (sic), . . .; unp. pap., archive of Ra'neferef, pi. 76 A 

• 3, ibd IV iht, sw 25; papyrus archive, temple of Neferirkare' 133 

• 5, ibd IV iht, . . . unp. pap., archive of Ra'neferef, pi. 76 C 

• 6, ibd IV prt, sw 22; wooden box for linen found in the tomb of Nefer 
and Kahay 134 

• 8, ibd IV smw; papyrus archive, mortuary temple of Neferirkare t135 

• 9; rock stela, Sinai 136 

• 10, ibd IV . . ., sw 24; papyrus archive, mortuary temple of Neferirkare' 137 
•11, ibd II iht, sw 11; papyrus archive, mortuary temple of Neferirkare' 138 

• 14, ibd tpy smw (ibd II smw); papyrus archive, mortuary temple of 
Neferirkare' 139 

132 The date, mentioned in Wepemneferet's will on a wall of his tomb (S. Hassan, 
Giza II, Cairo, 1936, fig. 219), was attributed by Spalinger ("Texts", 302, with a ref- 
erence to K. Baer, Rank and Title in the Old Kingdom, Chicago, 1960, 66) to Wenis. 
However, the persons mentioned in the tomb include a craftsman named Ra'neferef- 
ankh. If born in the reign of Ra'neferef, which seems probable, he could have been 
about 30 to 40 years old at the beginning of Djedkare"s reign. If this date be ascribed 
to Wenis, he would have been very old at the time of that king's acession. The date 
should, therefore, refer to Djedkare' rather than Wenis. 

133 Posener-Krieger & de Cenival, Abusir Papyri, pis. 13, 13 A; Posener-Krieger, Archives 
II, 480. 

134 H. Altenmuller & A. Moussa, The Tomb of Nefer and Kahay (Mainz: AV 5, 1971), 
18, 43-44, fig. 11. The excavators assigned the date to Djedkare'. However, Spalinger, 
"Texts", 302 suggested either Ra'neferef or Neuserre'. Surely Ra'neferef can be excluded 
(see above sub Ra'neferef). The tomb seems to have been built in the time of Neuserre', 
as the excavators surmised; see also, e.g., N. Cherpion, Mastabas et Hypogees de I'ancien 
Egypte (Bruxelles, 1989), 135. However, as pointed out by Altenmuller & Moussa, the 
burial in shaft no. 8, where the box with the date was found, was the last one made 
in the tomb and should be contemporary with Nefer's children. The dating of the 
inscription to the time of Djedkare is, therefore, very plausible. 

135 Posener-Krieger, de Cenival, Abusir Papyri, pis. 69, 69 A; Posener-Krieger, Archives 
II, 490. 

136 A. H. Gardiner, T. E. Peet & J. Cerny, The Inscriptions of Sinai I (London, 1952), 
pi. VIII, no. 14; II, 61. 

137 Posener-Krieger & de Cenival, Abusir Papyri, pis. 72, 72 A; Posener-Krieger, Archives 
II, 490. 

138 Posener-Krieger & de Cenival, Abusir Papyri, pis. 53, 53 A; Posener-Krieger, Archives 
II, 490. 

139 Posener-Krieger & de Cenival, Abusir Papyri, pis. 2, 2 A; Posener-Krieger, Archives 
II, 490. 


15, ibd IV prt, (sw) wpw; papyrus archive, mortuary temple of Nefe- 

rirkare t140 

15, ibd IV iht, sw 27; unp. pap., archive of Ra'neferef, pi. 20 B 

15, ibd IV iht, sw 28; unp. pap., archive of Ra'neferef 141 

15, ibd I, . . .; unp. pap., archive of Ra'neferef, pi. 21 L 

15 (n) T(nwt); unp. pap., archive of Ra'neferef, pi. 3 A 

16; papyrus archive, mortuary temple of Neferirkare c142 

[1]6, ibd IV smw, sw 28 w " 

17, ibd III; unp. pap., archive of Ra'neferef, pi. 8 D 

18, ibd III smw, sw c rq; unp. pap., archive of Ra'neferef, pi. 45 
18, ibd IV smw, sw c rq; unp. pap., archive of Ra'neferef, pi. 63 A 
21 (22 ?), ibd IV iht, sw 12; papyrus archive, mortuary temple of 
Neferirkare' 144 

damaged evidence of rnpt zp 

• 2 + x, ibd I . . ., sw . . .; unp. pap., archive of Ra'neferef, pi. 76 B 

• 10 + x; unp. pap., archive of Ra'neferef, pi. 85 C 

• 1 1 + x Tnw(t) ih ( c wt nb); unp. pap., archive of Ra'neferef, pi. 1 A. 

rnpt (m-)ht zp 

1, ibd IV iht, sw c rq; unp. pap., archive of Ra'neferef, pi. 77 A 

1, ibd IV smw, sw 1; unp. pap., archive of Ra'neferef, pi. 77 B 

tpy; unp. pap., archive of Ra'neferef, pi. 82 

3; rock stela, Sinai. 145 

4, ibd III smw, sw 15; unp. pap., archive of Ra'neferef, pi. 76 C. 

4, ibd III smw; unp. pap., archive of Ra'neferef, pi. 69 A 

7 (?), ibd I iht; unp. pap., archive of Ra'neferef, pi. 76 D 

10, ibd IV smw, sw 21; papyrus archive, mortuary temple of Neferir- 

kare c146 

140 Posener-Krieger & de Cenival, Abusir Papyri, pis. 47, 47 A; Posener-Krieger, Archives 
II, 490. 

141 Posener-Krieger, in: Melanges Mokhtar II (Cairo, 1985), 195-210. 

142 Posener-Krieger & de Cenival, Abusir Papyri, pis. 1, 1 A; Posener-Krieger, Archives 
II, 490. 

143 Urk I, 63.11; W. S. Smith, "Evidence", 113 n. 2; see also E. Eichler, SAK 18 
(1991), 146-147 (letter of Izezi to Senedjemib). 

144 Posener-Krieger & de Cenival, Abusir Papyri, pis. 41, 41 A; Posener-Krieger, Archives 
II, 490. Posener-Krieger transcribed the numeral following rnpt zp as 21. There is, 
however, the trace of another vertical stroke which allows the reconstruction '22'. 

145 Gardiner et al. (n. 136), I, pi. VII, no. 13; II, 60. 

14,1 Posener-Krieger & de Cenival, Abusir Papyri, pis. 14, 14 A; Posener-Krieger, Archives 
II, 490. 


• mpt zp (m)-xt zp (sic) 14, ibd I iht, sw 28; unp. pap., archive of Ra c ne- 
feref, pi. 66 B 

• 14, ibd . . .iht; unp. pap., archive of Ra'neferef, pi. 76 J 

• 17, ibd I smw (?), sw 23; sarcophagus of Idu, Abusir 147 

damaged evidence of mpt (m)-xt zp 

• 4 + x, ibd I iht; unp. pap., mortuary temple archive of Ra'neferef, 
pi. 77 I 

• 10 + x, . . .; unp. pap., mortuary temple archive of Ra'neferef, pi. 
76 E 

documents dating from the time of either Djedkare c or Wenis: 

• mpt zp 4, ibd IV prt, sw 2; papyrus archive, mortuary temple of 
Neferirkare cl4!i 

• mpt zp 4, ibd I smw, sw c rq; papyrus archive, mortuary temple of 
Neferirkare' 149 

• mpt zp 11, ibd III prt, sw 3 (?); mason's inscription, tomb of Rawer 


attested mpt zp: 1, 3, 4(?), 5, 6, 8, 9, 10, 11, 14, 15, 16, 17, 18, 21 


attested mpt (m)-xt zp: 1, 3, 4, 7 (?), 10, 14, 17 

mpt zp: mpt (m)-xt zp, 15(?): 7(?) 


TC 111.25: 30 years 

Manetho: Onnos — 33 years 

Beckerath, Chronologie, 155: 20 years 

147 M. Verner, SAKS (1980), 258-260, pi. 16. A new examination of the badly dam- 
aged inscription showed that the date should be read mpt (m-)ht zp 1 7 rather than mpt 
Zp 14, as suggested shortly after the discovery of the tomb. 

148 Posener-Krieger & de Cenival, Abusir Papyri, pis. 11, 11 A; Posener-Krieger, Archives 
II, 491, was hesitant about the attribution of this date to either Djedkare' or Wenis. 

149 Posener-Krieger & de Cenival, Abusir Papyri, pis. 11, 11 A; Posener-Krieger, Archives 
II, 491 was hesitant about the attribution of this date to either Djedkare' or Wenis. 

150 Junker, Giza III, 223-235; idem Giza VIII, 39f. Though a sealing bearing the 
name of Djedkare' was found in the tomb, the attribution of the date to him is some- 
what uncertain. Nevertheless Baer (n. 132), 98, assigned the tomb to the end of Dyn. 
5, while Harpur (n. 72), 213, dates it mid-Djedkare c to Wenis. 


rnpt zp 

• 3, ibd IV iht, sw 11; mason's inscription, anonymous mastaba "L", 
Abusir 1 ' 1 

• 6, ibd II smw, sw 28 

• 8, ibd IV smw; papyrus archive, mortuary temple of Neferirkare 1152 

rnpt (m)-xt zp 

• 4, ibd I prt 1 ™ 

Concerning several dates whose attribution is uncertain (Djedkare c or 
Wenis), see above under Djedkare c . 

attested rnpt zp'- 3, 6, 8 
attested rnpt (m)-xt zp'- 4 
rnpt zp-' rnpt (m)-xt zp, 3: 1 

Postscript. — Unfortunately, the edition of this volume has been delayed 
by four years. In the meantime, there has appeared some new infor- 
mation relating to the subject of my article to which I could not respond. 
For instance, a work by J. S. Nolan (The Original Lunar Calendar 
and Cattle Counts in Old Kingdom Egypt in: AH 17, 2003, 75-97) 
offering the explanation of the imbalance between the "years of an 
occasion" and "years after an occasion" in the contemporaneous Old 
Kingdom documents. Moreover, some new conclusions concerning the 
dated documents of the late Fifth Dynasty eventuated as a result of an 
examination of the papyri from Raneferef 's mortuary temple archive 
(see P. Posener-Krieger, M. Verner, H. Vymazalova, The Pyramid 
Complex of Raneferef. The Papyrus Archive, in press). It is thus a 
matter of some regret that this article could not be as comprehensive 
as I would have liked. 

151 M. Verner & V. Callender, Pfjedkave's Family Cemetery (Prague, 2002), 103. 

152 Posener-Krieger & de Cenival, Abusir Papyri, pis. 54 c, 54 A c; Posener-Krieger, 
Archives II, 491. 

153 Posener-Krieger & de Cenival, Abusir Papyri, pis. 50, 50 A; Posener-Krieger, Archives 
II, 491. 


Michel Baud 

Although data are quite abundant for Dyn. 6 and derive from a vari- 
ety of sources (royal annals and decrees on stone, administrative doc- 
uments on papyrus, expedition graffiti), the interpretation of the dating 
system used by the monarchy remains controversial. For this period, 
the dogma of the biennial census has been challenged in the most 
recent studies, especially in the compilations of Spalinger 1 and Helck 2 
prior to the publication of the South Saqqara Stone, with the royal 
annals of Dyn. 6. 3 In theory, the discovery of such a monument might 
be expected to clarify chronological questions, since the text is orga- 
nized in year-compartments ending with the citation of the year, either 
of the census type (mpt zp) or post-census type (rnpt m-ht zp)- But unfor- 
tunately, the inscriptions were quite systematically erased prior to the 
reuse of the slab as a sarcophagus lid, so that neither the demarcation 
of the compartments (which may in any case have been painted, rather 
than carved) nor most of the dates are preserved. Documents of 
significantly later date, such as the TC, do not provide any help for 
evaluating the dynasty's duration, or the lengths of individual reigns. 
Most, if not all, of the figures preserved are at odds with contempo- 
raneous OK data, despite repeated efforts to reconcile the two. 4 Therefore, 
the value of the papyrus lies more in the realm of historiography than 
in chronology.' 

1 Spalinger, "Texts", 275-319. 

2 W. Helck, "Gedanken zum Mord an Konig Teti", in: Fs Goedicke, 108-109. 

3 Baud & Dobrev, "Annales", 30-31; Baud & Dobrev, "Le verso des annales de la 
Vie dynastie, Pierre de Saqqara-Sud", BIFAO 97 (1997), 35-42. 

4 See Beckerath, Chronologic, 147-152. 

5 Baud, "Menes", 126-132. 


Reign of Teti 

No dates are preserved on the annals (recto, first register), 6 and the 
absence of lines dividing year-compartments does not even allow an 
estimate of the reign's total length. The space allotted Teti seems much 
too small to accommodate the available data; most likely, the monu- 
ment displayed a summary of the reign with compartments of very 
reduced size. Possibly what was initially considered the recto is rather 
the verso, 7 but arguments in favour of this are weak. 8 

The Abusir archive from the funerary temple of Neferirkare c pro- 
vides a number of dates for the early part of Teti's reign. 9 All come 
from pBerlin 10.474A-B recto and verso (HPBM V, 1969, pi. 92-95), 
a narrow roll which belongs to a single reign, as is obvious from the 
coherent time span of the recorded dates. Teti's serekh (pi. 94) clearly 
identifies the sovereign, at least for the left document of the presumed 
verso preserving the earliest dates: 

• [mpt (m)-ht zp 1], III prt sw c rk(y); rnpt (m)-ht zp 1 [III prt] sw 10; 
rnpt [(m)-ht zp 1] HI prt sw 9, in reverse order (pi. 94, left doc), 
to which the recto adds two: 

• rnpt (m)-ht zp 1, HI smw sw 3 (pi. 94, right doc), 

• rnpt (m)-ht zp [1], IV smw sw [x] and mpt zp 2, II smw sw 3, in 
normal order (pi. 92). 

pBerlin 15.729 verso (pi. 103) provides another date: [rnpt] (m)-ht zp 
[x], I >ht sw 3, but the year is missing and the identity of the king 
remains uncertain. It should not be Teti, 10 since the recto displays a 
basilophorous name citing a king Pepy 11 (not necessarily Pepy II). 
Whatever the numeral was, this document provides another example 
of a post-census year. 

6 Baud & Dobrev, "Annales", 23-92. 

7 V. Dobrev, "The South Saqqara Stone and the sarcophagus of Queen Mother 
Ankhesenpepy", in: M. Barta & J. Krejci, eds., Abusir and Saqqara in the Tear 2000 
(Prague: Archiv Orientalni Supplementa 9, 2001), 382-384. 

8 Baud & Dobrev, "Annales", 54; Baud & Dobrev (n. 3; 1997), 38. 

9 Posener-Krieger, Archives II, 491. 

10 Helck, Fs Goedicke. 

11 Posener-Krieger, Archives II, 491. 


Other dates for the reign of Teti are: 

• mpt (m)-ht zp 6, III smw sw [xj: a graffito at the alabaster quarry 
of Hatnub (gr. no. I); 12 

• mpt zp 11, I >ht sw 20: an ink inscription in the tomb of Nykau- 
Izezi (Teti Cemetery, Saqqara), added to a scene of fowling in the 
marshes, just above the boat in which the owner stands. 13 The 
inscription dates the official's burial "the 11th count, I >ht sw 20: 
burial in the necropolis of the prince, the treasurer of Lower Egypt, 
Nykau-Izezi". Although a king's name is not mentioned, all argu- 
ments favour the reign of Teti. By doubling the number of "occa- 
sions" hitherto known, this new date raises several questions (see 

Reign of Userkare' 

The South Saqqara royal annals, 14 demonstrate conclusively the exis- 
tence of this king, but almost nothing remains of the section in the 
middle of the uppermost register devoted to his reign. The available 
space between the titularies of Teti and Pepy I, when compared to the 
size of an average year compartment of the latter, indicates that 
Userkare°s reign must have been brief, from two to four years. This 
conclusion is consistent with the very few monuments of this king, 
mostly seal impressions, so far recovered. The silence of contempora- 
neous private biographies is disturbing. A funerary complex planned 
but never erected, is a better explanation for this absence than a spec- 
ulative damnatio memoriae.^ Regardless, the chronographic purpose of the 
royal annals did not allow the omission of this king, whatever form his 
titulary may have taken. 

12 Eichler, E., Untersuchungen zum Expeditionswesen des agyptischen Altai Reiches (Wiesbaden, 
1993), 41, no. 36. 

13 (a) N. Kanawati & M. Abder-Raziq, The Teti Cemetery at Saqqara VI. The Tomb of 
Mkauisesi (Warminster: ACE Reports 14, 2000), pi. 50; (b) N. Kanawati, "A new hit/ 
rnpt-zp for Teti and its implication for Old Kingdom chronology", GM 177 (2000) 

14 Baud & Dobrev, "Annales", 28, 53, 59-62. 

15 R. Stadelmann, "Konig Teti und der Beginn der 6. Dynastie", in: Fs Leclant I, 


Reign of Pepy I 

Again, the discussion must start with the data preserved in the annals 
stone from South Saqqara. The reign extends from the last third of 
the first register (A) to the very beginning of the fifth (E). Although 
there is again no demarcation of year-compartments, traces of a num- 
ber of memorial formulae (nswt bjt Ppy jrn.f m mnw.f) provide clues for 
reconstructing the original layout. 16 The twelve surviving formulae 
(M3— Ml 4) are spaced at rather regular intervals (x 2 or X 3 where 
one, or perhaps two formulae are lacking), which supports as estimate 
of the original number at up to 25. Since both "occasion" and "after- 
occasion" years are known for the reign, obviously each mww-formula 
was associated with a pair of years, a census year and a post-census 
year, presuming a regular biennial system. 

Contrary to the editio princeps of the monument, 17 it is, however, by 
no means certain that a single heading systematically covered two years. 
Some compartments, especially at the beginning of a register, indeed 
appear much larger than others (see especially M5, second reg., and 
one formula before M10, fourth reg.). Therefore, it is tempting to con- 
clude that they group two different years, by contrast to the others, 
which represent the vast majority of the (theoretical) compartments. But 
there are a number of objections to such a radical proposition. First, 
the size of a compartment might vary slightly according to the num- 
ber of available and/or relevant data that needed to be recorded for 
posterity, as exemplified by the difference between the first and the sec- 
ond register where the average distance between successive mnw is 
respectively 30 and 40 cm. For years of crucial importance to the 
monarchy, such as the royal jubilee, the compartments could have been 
much larger than the average, although this is not an absolute neces- 
sity (cf. the minimal size of the first year, dedicated to the coronation 
rites). Secondly, there is at least one surviving example of two years 
grouped in a (theoretical) compartment of average size, reg. D, for- 
mula following Mil (hereafter Mil + 1). The date preserved here, a 
census year, occupies the middle of the available space and not the 
end, as anticipated for the case of a single year covered by its own 
mnw-formula. Nonetheless, if it is possible that most of the memorial 

Baud & Dobrev, "Annales", 50—53. 

Baud & Dobrev, "Annales", 50-52, fig. 19. 


formulae grouped census and post-census years together, the present 
condition of the stone leaves some doubt about the generalization of 
such a layout. Only parallel evidence from other sources might help 
to solve this problem. The dates preserved in the annals are indeed 
very few for the reign of Pepy I, and not unproblematic as regards 
their reading: 

• rnpt zp 18 (reg. D, Mil + 1, text zone D4); 

• rnpt (m)-ht zp 23 (reg. E, M14 + 1, text zone E7); 

• rnpt [m-ht ?] zp 25 (reg. E, Ml 4 + 3, text zone E8). 

These high counts are also attested in several expedition graffiti 18 and 
a royal decree: 19 

• rnpt (m)-ht zp 18, III smw sw 27: Wadi Hammamat graffito no. 107, 
mentioning the first jubilee; 

• rnpt m-ht zp 18, IV smw sw 5: Sinai graffito no. 16, mentioning the 
first jubilee; 

• rnpt zp 21, I prt sw 23: decree for the Pyramid complex of Snofru, 
Dahshur; 20 

• rnpt zp 25, I >ht sw [xj: Hatnub quarry graffito no. Ill, once more 
associated with the first jubilee. 

Major clearance work at the king's pyramid, South Saqqara, 1987—88 
and 1993—97, revealed a few dates among the great number of masons' 
marks. 21 Most did not include the year but, according to common prac- 
tice, 22 only a season, month and day. 23 A block from the eastern end 
of the south side is a notable exception. 24 After the group rnpt(?)-zp 
there is an hieratic sign, which at first sight reads 30, followed by two 
vertical strokes. Such a high date, count 32 (or even 22), from an early 


18 Eichler (n. 12), nos. 133, 16, 30. 

19 Spalinger, "Texts", 303-304. 
Goedicke, Dokumente, 55-77, fig. 5. 
A. Labrousse, I 'architecture des pyramides a textes. II, Saqqara Sud, (Cairo: BdE 131, 

2000), 1-2. 

22 For the mastaba of Ptahshepses see M. Verner, Abusir II. Baugraffiti der Ptahschepses- 
Mastaba (Prague, 1992), 176-198. 

23 V. Dobrev, "Observations sur quelques marques de la pyramide Pepi Ier", in: Fs 
leclant I, 150-151. 

24 Dobrev, pers. comm. 1994; see also F. Raffaele in 2001, "Sixth Dynasty Annals. 
The South Saqqara Stone", 

The mark is visible in Labrousse (n. 21), fig. 10, but at a very reduced scale. 


stage of the construction (corresponding to the 5th course of revetment 
blocks), leaves some doubt about this reading. Furthermore, since the 
highest count of the reign is 25, as confirmed by both the royal annals 
and the other contemporary data, the only satisfactory solution is to 
suppose that a double system of counting, one annual, the other bien- 
nial (whether regular or not) existed. The annual rnpt-zp 32 should be 
therefore equated with a lesser number of biennial census counts, at 
least 1 6 if regular. (Note that an anomalous group of 1 9 strokes in two 
rows (10 + 9) just before the date is obviously a calculation of some 
sort, not necessarily connected to the dating system). At present, and 
until adequate publication, this intriguing problem defies solution. 
Conceivably, it may eventually provide a key to explaining the con- 
tradictory dates for the king's first jubilee. 

The date of Pepy's first heb-sed is controversial, since it is associ- 
ated with two different years, counts 18+ and 25. 2s Taking the two 
figures at face value, they would have been separated by at least 6 
years (systematically excluding post-census dates in this period, which 
is unlikely), and as many as 12 (with systematic post-census years). 
Therefore, the significance of both or either might reasonably be chal- 
lenged, and could attest the prevalence of the Wunsch-Idee in the men- 
tion of the jubilee, for the benefit of the king's longevity. 26 A strictly 
historical/chronological interpretation is, however, still possible. Spalinger 
ingeniously envisaged the existence of two parallel dating systems at 
this period, one annual, i.e., 25 counts, the other (irregularly) biennial, 
i.e., 18 census counts, plus presumably 7 post-census. 27 Although inter- 
pretation of building graffiti may support this hypothesis, there remains 
the very confusing consequences of such a theoretical double system, 
both citing all years simply rnpt-zp. Furthermore, Spalinger has not con- 
sidered one important factor: the context in which the dates are actu- 
ally associated with the jubilee. As for the Sinai relief, there is no direct 
equation between the first jubilee and the date of the expedition. And 
though the scene depicts the royal ceremony in a format typical of an 

25 These do not refer to two different jubilees as P. O'Mara ("Dating the Sed- 
Festival: Was there a Single Model?", GM 136 [1993], 57-70) thought, nor can the 
second belong to Pepy II (as proposed by J. v. Beckerath, "Gedanken zu den Daten 
der Sed-Feste", MDAIK 47 [1991], 30; tentatively Eichler [n. 12], 39). 

26 E. Hornung, "Sedfest und Geschichte", MDAIK 47 (1991), 169-171; with earlier 

27 Spalinger, "Texts", 305-306. 


year compartment in the annals, it is an all too frequent royal tableau 
to be taken as a true date. 28 

The same may possibly hold true for the other inscriptions, although 
the same historical connection between the Sinai and the Hammamat 
graffiti, both under count 18+, end of the shemu-season, may not be 
fortuitous. 29 However, a tendency to mention the jubilee repeatedly in 
the years following its celebration apparently existed, 30 for example, in 
connection with intense building activity at the royal funerary complex, 
down until the very end of the reign (i.e., count 25). Here again, the 
royal annals furnish a new argument favouring this hypothesis. Between 
the mention of count 18 and the next memorial formula which belongs 
to count 19, end of register D, the available space for count 18+ is 
the expected half of the average size of a theoretical compartment. It 
is hard to believe that such a narrow space corresponds to the jubilee 
celebration, which obviously had a considerable importance for this 
(and every) king, as is documented, for example, by the number of 
stone vessels celebrating the event. 31 (Count 25, the very last compart- 
ment of the annals, is of course excepted). By contrast, the longest com- 
partment of the reign — more than half again the average length — is 
M10— 1 (i.e., one formula before the preserved M10) at the beginning 
of register D. Fortuitously or not, this compartment corresponds pre- 
cisely to year 30/31, if a strictly biennial system of numbering is pre- 
sumed. 32 This could also explain why the handful of documents dated 
to the first jubilee did not cite any other date. For example, decree 
Coptos A simply epitomized the rule for the renewal of the king's pow- 
ers after 30 years. 33 

Specialists, however, remain divided on whether this rule obtained 
during the OK. 34 There would therefore be no necessity to place the 

28 See, too, two inscriptions recording Merenre"s visit to the First Cataract area to 
receive the hommage of Nubian chief(s). One displays a real date (count 5, see infra), 
reign of Merenre'; the other only a pictorial zema-tavoy (JJrk. I, 1 1 1), which may be 
indicative of the theoretical date — the coronation year — in which such an event would 
have taken place. 

29 J. Vercoutter, L'Egypte et la vallee du Nil, 1: De.s engines a la fin de I'Ancien Empire 
(Paris, 1992), 326. 

30 See Hornung (n. 26), 170. 

31 A. Minault-Gout, "Sur les vases jubilaires et leur diffusion", in C. Berger & 
B. Mathieu, eds., Fs Jean-Philippe Lauer (Montpellier, 1997), 305-14. 

32 Raffaele (n. 24). 

:i:i Goedicke, Dokumente, 41-54, fig. 4. 

34 For the range of interpretations see Beckerath (n. 25), Hornung (n. 26), O'Mara 
(n. 25), A. A. Krol, "The representation of the 'Sed-Platform' in the Early Dynastic 
monuments", GM 184 (2001), 27-37. 


jubilee as late as year 35/36, 35 nor to equate rnpt zp 18+, presuming 
Userkare c was a usurper, with a canonical year 30 of strictly personal 
rule, 36 an assumption invalidated by the royal annals. 37 

Reign of Merenre c 

The last register (F) of the recto of the Dyn. 6 annals is dedicated to 
the first years of this reign. The number of compartments is uncertain, 
but five or six is a reasonable estimate. 38 The dates preserved are: 

• rnpt zmi-Swy, associated with the first memorial formula (Ml 5) of 
the register (text zone Fl); 

• rnpt zp 1 (+ 1?), with the second formula (Ml 6, zone F3); 

• mpt (m)-ht zp 1 (+ 1?), probably with the next formula, not pre- 
served (zone F5). 

In our initial publication, we logically assumed that the two last dates 
were to be read as counts 2 and 2+, since in these annals, the Unification 
of the Two Lands was apparently considered a year of cattle census 
(the expression tnwtjh does follow the date). After this first census (count 1), 
a compartment was tentatively delineated to account for a post-census 
year after the Unification, 39 considering that for such years also the sys- 
tem remained biennial. Although this remains a possibility, there are 
weaknesses in such a reconstruction. In the first place, this so-called 
count 1+ would be confined to a very narrow space, when compared 
to the other very broad compartments of the last register. 40 Secondly, 
the figure of the next date consists of a very deeply carved single 
stroke, 41 and it is unlikely that another stroke ever existed; there is also 
no space available for an alleged second stroke under the zp sign. It 
may not be mere chance that the next date also retains only a stroke. 
These two dates should be read accordingly as counts 1 and 1+, even 
if the first year of the reign was labelled census year. Should the suc- 
ceeding years be read 1/1+ or 2/2+, it is nonetheless clear that this 

35 Contra e.g., H. Goedicke, "Two Mining Records from the Wadi Hammamat" 
RdE 41 (1990), 65-93, at 67, and O'Mara (n. 25). 

36 Contra Spalinger, "Texts", 305-306. 

37 Baud & Dobrev, "Annales", 61-62. 

38 Baud & Dobrev, "Annales", 48-49, 54. 

39 Baud & Dobrev, "Annales", fig. 19. 

40 Baud, "Menes", 123-124. 

41 Baud & Dobrev, "Annales", pi. VII c. 


period experienced a biennial census. Since a year 5+ is also known 
(see below) and since it is likely that the document did not end abruptly 
in mid-reign, it must be concluded that the reign continued on the 
verso (see above, contra Dobrev), 42 even if the titulary of this king prob- 
ably featured in the introductory column of the recto, as recently pro- 
posed, citing new and convincing arguments. 43 Other sources are restricted 
to two rock inscriptions: 44 

• rnpt zp 5, II smw sw 28, First Cataract area, Urk. I, 110, 12; 

• rnpt (m)-ht 5, Hatnub graffito no. VI. 45 

Once again, it is clear that the latest years of the reign experienced a 
biennial system. 

Reign of Pepy II 

Spalinger's list 46 remains relatively current and must be consulted for 
further references: 

• rnpt (m)-ht zp tpy, TV >ht sw 10 (in two parts) and rnpt zp 2, pro- 
cession graffiti in Wadi Hilal (El Kab), although the beginning of 
the reign of Pepy II is a good possibility, 4 ' the dates could belong 
to his predecessor Merenre c ; 

• rnpt zp 2, III 3ht sw 15, letter of the king to Harkhuf in his tomb, 

• rnpt zp 2, Sinai graffito no. 17; 4!i 

• rnpt zp 11, I smw sw 23, the famous letter found in the workshop 
adjacent to Temple T in the Djoser complex, Saqqara; the reign 
is inferred from other chronological data of the archives; 49 

• rnpt (m)-ht zp H, II smw sw 26, decree Coptos B, temple of Min; 

42 Dobrev (n. 7). 

43 Dobrev (n. 7), 384-385, pi. 58. 

44 Spalinger, "Texts", 306-307. 

45 Eichler (n. 12), 40, no. 33. 

46 Spalinger, "Texts", 307-308. 

47 H. Vandekerckhove & R. Miiller-Wollermann, Elkab VI. Die Felsinschriften des Wadi 
HUM (Turnhout, 2001), 210-211, gr. O 144, and 183-186, gr. O 74; conclusions: 

48 Eichler (n. 12) 35, no. 17. 

19 P. Posener-Krieger, "Fragments de papyrus provenant de Saqqarah", RdE 32 
(1980), 83-93. 


• mpt zp 12, graffito of Tomas, either of the two Pepys, but more 
probably Pepy II; 50 

• mpt zp 14, I >ht sw 23(7), Hatnub graffito no. 3; 51 

• mpt (m)-ht 22, IV smw sw 28 (date in two parts), decree Coptos C; 

• mpt zp 31, III >ht sw 3 [+ 3], decree for the cult of Mycerinus, 

• mpt zp 31(P), IV prt sw [x], graffito from the king's funerary temple, 

• mpt (m)-ht zp 31, I smw sw 20, Hatnub graffito no. 7;' 2 

• mpt zp 33 (?) or 24 (?), IV sw [x], decree for the cult of Queen 
Udjebten, Saqqara. 

There are also dates without a king's name which can be placed securely 
in the second half of Dyn. 6. The first two are from Giza:' 3 

• mpt zp 2, III prt sw 27, two mason's marks on the walls of mastaba 
G 7803C, Giza Eastern Cemetery, and another citing the same 
year, but month, season and day lost; 

• mpt (m)-ht zp 5, III prt sw 29, two execration texts from Giza. 
Another date comes from an expedition graffito at Tomas;' 4 

• mpt 6, III smw, probably either Pepy I or II. 

Biennial, Irregular or Annual Census? The Case of Dynasty 6 

The regularity of the census, backbone of the Ancient Egyptian dating 
system, is still a matter of controversy for the OK. The most recent 
discussions of this crucial problem present the largest possible spectrum 
of interpretations, ranging from a regular biennial census" through an 
annual census with post-census years at irregular intervals' 6 to a strictiy 
annual one. 57 That an annual count was already established by Dyn. 

50 According to Eichler (n. 12), 105 (no. 227A). 

51 Eichler (n. 12), 43, no. 39. 

52 Eichler (n. 12), 44-45, no. 43. 

53 Spalinger, "Texts", 308-309. 

54 Eichler (n. 12), 109, no. 245. 

55 E.g., Baud, "Menes". 

56 E.g., Verner, "Archaeological Remarks on the 4th and 5th Dynasty Chronology", 
Archiv Orientdlni 69 (2001), 410-412. 

57 E.g., Kanawati (n. 13). 


6 (and not in the FIP, according to the traditional view), 58 is disproved 
by the number of attestations of m-ht zp years at this period. Kanawati 
believes, however, that they resulted from provisional numbering, sub- 
sequently altered to "normal 2 years counts". For example, "the ref- 
erence to the 'y ear after the sixth count' may simply refer to the seventh 
year, but before the seventh count was undertaken". 39 This hypothesis 
ignores the existence of the South Saqqara Stone, with at least two 
examples of post-census years (one in the reign of Pepy I, and the sec- 
ond under Merenre c , see above). Since annals are an official recapitu- 
lation of events, there is no reason why the entries should reflect a 
provisional numbering system. 

Kanawati's proposal is an attempt to reconcile apparently contra- 
dictory data in the case of Nykau-Izezi (see above), viz., (a) a basilophorous 
name suggesting that Nykau-Izezi was born under Izezi 60 (b) his rep- 
resentation in the reliefs of the causeway of Wenis, with the high rank- 
ing title 'sole friend'; (c) the dating of his burial to the 11th count, 
presumably of Teti; (d) an estimate of his age at death, based on exam- 
ination of his remains, as 40—45 years or even slightly younger (35). 

Kanawati was influenced by the difficulty of reconciling the relative 
brevity of Nykau-Izezi's life with the time-span between Djedkare°s 
reign and the 11th census in Teti's reign, presuming a regular bien- 
nial census. The 11th census of Teti corresponds to year 22/23 of a 
regular biennial census system, but at least 13/14, if the census was 
irregular (since two intervening years are known, 1+ and 6+, see above). 
Nykau-Izezi was therefore between about 17/18 and 26/27 years old 
when Teti ascended the throne. Since, on the same premise, Wenis 
reigned between 16 (rnpt zp 8 as highest census) and 9 years, 61 the 
official was either born at the very end of Djedkare°s reign, or ten 
years earlier. This would account for Izezi in his name, but this expla- 
nation is superfluous, since kings were celebrated thus for various rea- 
sons, if indeed such names were not simply passed from father to son. 62 
Thus the name does not prove that Nykau-Izezi's career began in 
Djedkare°s reign. If he is the like-named official in the Wenis cause- 

58 E.g., Gardiner, "Years", 14-16. 

59 Kanawati (n. 13; 2000a), 21, 23, b; see also Helck (n. 2), 110. 

60 N. Kanawati & M. Abder-Raziq, The Teti Cemetery at Saqqara V. The Tomb of Hesi 
(Warminster: ACE Reports 13, 1999), 37-38, pi. 33, 59. 

61 Verner (n. 56), 410-412, 416. 

62 Another Nykau-Izezi is mentioned, for example, on three graffiti at the pyramid 
of Pepy I, see V. Dobrev, "Les marques sur pierres de construction de la necropole 
de Pepi Ier. Etude prosopographique", BIFAO 96 (1996), 112, D.l. 


way reliefs, he was promoted to 'sole friend' between the ages of 17 
and 27 — quite young in either case. The first alternative may indeed 
seem much too young, but high-ranking titles may not have been 
confined to mature officials. All in all, there are simply too many uncer- 
tainties in Kanawati's argument. The same situation obtains for other 
officials who started their careers under Teti and died under Merenre'. 
Some cases may still be debatable, as Kanawati rightly pointed out. 63 
Weni, for example, already held a relatively high position under Teti 
and must have been about 70 when Merenre c ascended the throne, 
since Pepy I's reign amounts to 50 years, presuming a strictly biennial 
system. This seems very old for the onerous duties reported in his biog- 
raphy, and for his journeys to obtain materials and monuments for the 
king's tomb. However, this may have been a conceit to celebrate the 
official's longevity and his capacity to remain active at an advanced age. 
If the theory of an annual census be discarded for sound reasons, it 
must nevertheless be admitted that no basis exists for deciding in favour 
of either of the alternatives, the regular or irregular biennial system. 
On the one hand, the number of attested census years is very well bal- 
anced by post-census years during the reigns of Teti to Pepy I. When 
the sources shed some light on a segment of these reigns, i.e. the ear- 
liest counts of Teti (1 to 2), the latest counts of Pepy I (18 to 25) and 
the earliest and latest counts of Merenre c (1 and 5), the number of 
intervening years equals, or nearly equals, the number of census years. 
The alleged total imbalance between the two 64 results from the limited 
number of sources and prejudice associated with the category of sources, 
as exemplified by the masons' marks of Dyn. 4 at Giza. 65 On the other 
hand, the extreme imbalance for the reign of Pepy II could favour an 
irregular counting system in his particular case. The celebrated longevity 
of the king in tradition 66 as well as the fact that he was a child at his 
accession, 67 demand reconciliation with the contemporaneous record, 
to which a biennial count does justice. 68 The table below summarizes 
the dates for the period from Teti to Pepy II: 

63 Kanawati (n. 13, 2000a), 22-23. 

64 E.g., Helck (n. 2), 106-110; Spalinger, "Texts", 314-316. 

65 Baud, "Menes", 119-121. 

66 Baud, "Menes", 129, with caution. 

'" K. Ryholt, "The Late Old Kingdom in the Turin King-list and the Identity of 
Nitocris", ZAS 127 (2000), 87-100, at 94. 

68 H. Goedicke, "The Death of Pepy II-Neferkare", SAK 15 (1988), 111-121; Beckerath, 
Chronologie, 151-152. 





Minimal Number 
of Post-census 

Minimal Reign 
Length A 

Mnimal Reign 
Length B 


Pepy I 
Pepy II 




31, ev. 33 


2 or 3 

3 or 4 

13 years 




22/23 years 






min. 83 




X+ indicates a post-census year, for which X is the number of counts; 
"minimal reign length A" is the sum of the highest count (col. 2) and 
the attested intervening years (col. 3); 69 "minimal reign length B" pre- 
supposes a regular biennial census; the estimate for Userkare c is based 
on the royal annals (see above). 

Dynasty 8 

The identity, number and order of the Memphite rulers of Dyn. 8 
remain uncertain and identifications rely heavily on the much later 
Ramesside lists. 70 The TC counts 8 rulers after Pepy II (col. iv, no. 5 
to 13, this name and some others in lacuna).' 1 It includes, in second 
position, Queen Nitocris, who turns out to be a male ruler, Neitiqerty 
Siptah, according to Ryholt's recent examination of the papyrus. 72 The 
Abydos list (nos. 39 to 56) adds 10 more rulers, all probably between 
Neitiqerty (Abydos no. 40, if identified with Netjerkare c ) and Neferka 
Khered-seneb (Abydos no. 51, called Neferkare' Pepy-seneb),' 3 a group 
which may have been in lacuna in the TC Vorlage and therefore prob- 
ably reported as wsf/\ost. n 

69 After Vemer (n. 56), 415-416. 

70 E.g., Beckerath, Chronologie, 151-152. 

" This column should be renumbered 5, since according to Ryholt's recent study 
(see n. 67) there is evidence of an intermediate column between col. I and II of 
Gardiner's edition. 

72 Ryholt (n. 67), 87-100. 

73 See Ryholt (n. 67), 87-94. 

74 Beckerath, Chronologie, 148-9; Ryholt (n. 67), 96-98. 


According to the partially preserved figures in the TC, both for reign 
lengths and summations of grouped dynasties, Dyn. 8 covered a very 
short period of about one generation. However, not only is the read- 
ing of some of the figures problematic (e.g., the total for Merenre'), 75 
but the very value of the figures remains largely questionable, as 
exemplified by contradictory OK data for a number of reigns. Recent 
analysis would double the duration of this dynasty, to at least 50 years, 76 
or even slightly more, 7 ' but this is not a significant change in the image 
of a relatively short and obscure period. In this particular case, the TC 
data is probably not far from the truth with its low figures for indi- 
vidual reigns: 1 year for the immediate successor of Pepy II (name lost, 
no. 6) and between 1 1/2 to 4 years for the last four rulers (nos. 
10-13). The six wsj-years reported in the subtotals (col. iv, 14-17) for 
ten missing kings probably represent an artificial emendation of the 
scribe, as exemplified by other occurrences of this figure. 78 All in all, 
these brief reigns accord with the few royal monuments recovered so 
far, and the low figures of the preserved dates. 79 Arranged in increas- 
ing numeric order, they are: 

• mpt zrri>-8wy, II prt sw 20, Coptos decree P of [Netjeri-bau] (Horus 
name of Neferkauhor; identification from parallel decrees), 80 tem- 
ple of Min; 81 

• mpt zp zmS-Bwy, IV smw J (wpty), decree of [Demedj-ib]-tawy (?), 
(Horus name), 82 funerary complex of Queen Neith, Saqqara; 

75 For the old reading '44 years', see Ryholt (n. 67), 90, 98. 

76 Beckerath, Chronologie, 151-152. 

" S. Seidlmayer, "Zwei Anmerkungen zur Dynastie der Herakleopoliten", GM 157 
(1997), 84-85. 

78 Ryholt (n. 67), 97-98. 

79 Spalinger, "Texts", 312-313. — Note also the inscription Cairo JE 43290 dated to 
mpt zp <1 ?>; IV >ht 25. The numeral is omitted, but 1 is the most likely emenda- 
tion, see H. Goedicke, "A Cult Inventory of the Eighth Dynasty from Coptos (Cairo 
JE 43290)", MDAIK 50 (1994), 72. This could refer to the first incomplete civil year- 
year — usually designated zm>-$wy. The inscription presumably originates from Coptos 
or nearby Khozam. Goedicke (ibidem) ascribed it tentatively to Nefer-kau-hor, but 
Fischer, in: Manuelian, ed., Studies Simpson, 267-270, argues for a date towards the end 
of Herakleopolitan rule in the Coptite nome. See also below Krauss, chapter III. 8 
for the Khozam lunar date. 

80 See W. C. Hayes, "Royal Decrees from the Temple of Min at Coptos", JEA 32 
(1946), 3-23. 

81 Goedicke, Dokumente, 195-196, with Hayes (n. 80), pi. V. 

82 According to the restoration proposed by Schenkel, Memphis, 24-25. 


• rnpt zp tpy, IV >ht sw 2, Wadi Hammamat inscription of King Ity 
(mentioned in the name of his presumed pyramid), 83 possibly Dyn. 
8 (O.Ham no. 169); 

• rnpt zp tpy (?), Ill smw sw 2, Wadi Hammamat inscription of an 
unknown king, date uncertain but possibly Dyn. 8 (O.Ham no. 152); K4 

• [rnpt] zp 4 [+ x?J, n5 season etc. lost, Coptos decree H of king 
Kha[bau?] (Horus name). 86 

The absence of post-census years probably testifies to a change in the 
dating system from a regular (?) biennial to an annual one. 


8:i Tentatively equated with Neferirkare' II of the Abydos list and the contempo- 
rary Horus Demedjibtawy by Spalinger, "Texts", 313, and n. 104. Goedicke's read- 
ing (n. 35), 66-67, (rnpt zp tpy <jb-sd> taken to refer to Pepy I is not supported by 
the parallel evidence. 

84 See Schenkel, Memphis, 32-33; Goedicke (n. 79), 83. 

85 The stela is lost below the four aligned strokes. While 3 or even 4 more strokes 
could possibly have figured in a lower line, it is rather unlikely in such a period of 
ephemeral kings. 

86 Goedicke, Dokumente, 163-164, fig. 16 and 23; for the date: Hayes (n. 80), 13, 
n. 7, and pi. iiia, top, before col. 1. 

87 Gardiner, "Years", 14-16; Hayes (n. 80), 13; Spalinger, "Texts", 312. 


Stephan J. Seidlmayer 

In dynastic history, 1 the FIP spans the era of Herakleopolitan rule 
(Dyns. 9 and 10) and the earlier part of the Theban Dyn. 11 up to 
the re-unification of the country which occurred at some point in the 
reign of Nebhepetre c Mentuhotpe II. When we attempt to clarify the 
dynastic structure of this period and to estimate its chronological length, 
our argument will be more straightforward if we first deal with Dyn. 
1 1 and only then turn our attention to the problems of the Herakleo- 
politan dynasties. 

Dyn. 11 

We are comparatively well informed about Dyn. II. 2 While the king 
lists of Abydos and Saqqara omitted all FIP rulers and listed only 
Nebhepetre c Mentuhotpe II and S c ankhkare c Mentuhotpe III of Dyn. 
II, 3 the TC gives a full account of its rulers, omitting only (as do the 
lists of Abydos and Saqqara) its last ruler Nebtawyre c Mentuhotpe IV, 
whose reign was relegated to a group of "missing" (wsf) years. 4 While 
the names of most of the kings are destroyed in the TC, a number of 
reign length data are preserved. Most valuable, however, is the fact 
that this document also provides a figure for the total length of the 
dynasty which allows us to determine the combined length of the first 
two reigns, for which individual length data are not preserved. The 
royal names which are destroyed in the TC can easily be reconstructed. 

1 To define the FIP as a distinctive phase in the history of pharaonic culture or its 
political structure, one would envisage a more extensive period including at least the 
end of the OK after the demise of Pepy II. 

2 For accounts of the chronology of Dyn. 11 see Schenkel, Studim, 145—149; Gestermann, 
Kontinuitdt, 22—31; Beckerath, Chronologie, 139-142. 

3 Abydos list nos. 57-58, Saqqara list nos. 37-38. 

4 TCV, 11-18. 



The list of royal ancestors inscribed by Tuthmosis III in the temple of 
Karnak' and a relief block from the temple of Tod which lists the pre- 
decessors of king Mentuhotpe II are particularly relevant. 6 Biographical 
inscriptions of officials also attest the names and sequence of the more 
important rulers, 7 while dated monuments and inscriptions concur to 
confirm the accuracy of the reign length data of the TC. 8 Confusion 
about the number of rulers named Mentuhotpe, caused by the fact 
that Nebhepetre c Mentuhotpe II changed his protocol twice, were set- 
tled by Gardiner. 9 On this basis, the data relating to the sequence and 
length of reigns of Dyn. 1 1 can be summarized with some confidence 
as follows: 




Urk. IV 



jrj-p't [hUj-'J Jnj-fjtj.f] 608.15 count Inyotef 

[Mntw-htp] 16 Hrw tpj-' Mn[tw-htp] 608.14 Mentuhotpe I 

[Jnj-jtjf] Hrw [S]h[r-t'.wj] Jnj-[jtjf] 608.13 Hrw Shr-i'wj 

Jnj-jtjf Inyotef I 
[JMJ-jtj.f] 49 Hrw [Wlh-'nh] Jnj-jtjf 608.12 Hrw Wh-'nh 

Jnj-jtjf Inyotef II 
[Jnj-jtj.f] 8 [ [Hrw JVht-nb-tp 

nfrj Jnj-jtjf Inyotef III 

Nb-hpt-R 51 Nb-hpt-R 609.14 Mentuhotpe II 

S'nh-k,'-R 12 S'nh-ki-R 609.15 Mentuhotpe III 

wsf 7 [Nh-]t',[wj]-R 609.16 Mentuhotpe IV 

Total 143 

5 Urk. IV, 608-609. 

6 J. Vandier, "Un nouvel roi Antef de la XF dynastie", BIFAO 36 (1936), 101-116. 

7 The sequence Inyotef II, Inyotef III, Mentuhotpe II is attested in three biographical 
inscriptions from Thebes, see Clere & Vandier, Textes, 15-16 § 20, and 19-20 § 23-24. 

8 For Inyotef II a regnal year 50, probably the year of his burial, is attested on a 
stela from his tomb, Clere & Vandier, Textes, 11 § 1 6; no dated monuments are known 
for Inyotef III; for Mentuhotpe II, stela Turin 1447 (Schenkel, Memphis, 240) attests 
year 46; for Mentuhotpe III a rock inscription in Wadi Hammamat attests year 8 
(Schenkel, Memphis, 253 no. 426), and for Mentuhotpe IV a series of graffiti at the 
same place, year 2 (Schenkel, Memphis, 263-268, nos. 441-444). 

9 A. H. Gardiner, "The First King Menthotpe of the Eleventh Dynasty", MDAIK 
14 (1956), 42-51; see also L. Habachi, "King Nebhepetre Menthuhotep: his monu- 
ments, place in history, deification and unusual representations in the form of gods", 
MDAIK 19 (1963), 16-52. 


A number of comments are in order. "Count" Inyotef, who was prefixed 
as a non-royal ancestor to the line of Dyn. 1 1 kings in the Karnak 
inscriptions, is in all probability identical with the jrj-p'-t h>.tj- c Jnj-jtj.f 
ms Jkwj to whom Senwosret I dedicated a votive statue in this temple. 10 
Whether he can also be identified with one of the attested pre- 11th 
dynasty nomarchs of this name," cannot be strictly demonstrated. The 
lacunary state of the evidence available would seem to recommend a 
careful stance in such matters. In any case, this question has no direct 
consequences for the chronology of the period. 

The Horus name tpj- e "the ancestor" accorded the first king of the 
dynasty is evidently a later fiction intended to prolong the royal line 
into the past. No contemporary monuments are attested for this person; 
however, a statue erected by Inyotef II at Elephantine calls him "father 
of the gods", i.e. the father of the first two kings of Dyn. II. 12 Nor is 
his direct successor, Sehertawy Inyotef I ever attested in contemporary 
inscriptions. 13 However, the sequence of tombs in the royal necropolis 
at el-Tarif suggests that the Saff Dawaba belonged to this ruler. The 
truly extraordinary size and layout of this tomb substantiate his claim 
to royal status. 

All other kings of the dynasty are well documented in contemporary 
sources. Of some interest is the fact that the last ruler, Nebtawyre c 
Mentuhotpe IV, attested in rock inscriptions from his second regnal 
year in Wadi Hammamat, 14 was omitted from all of the NK kings lists. 
The reason is speculative, but it does not seem very far fetched to sup- 
pose that his absence was motivated in some way by the circumstances 
of the transfer of power to a new royal house. In view of the lack of 
sufficient evidence it remains uncertain whether all 7 "missing"-years 
in the TC belonged to him or whether there was a period of disputed 
rule at the end of the dynasty, 1 ' for which, however, there is no pos- 
itive evidence. 

10 Cairo CG 42005. 

11 See the discussion in Gestermann, Kontinuitat, 24-26, and F. Gomaa, Agypten wahrend 
der Ersten Zmschenzeit (Wiesbaden: TAVO B27, 1980), 138-144. 

12 On this king see L. Habachi, "God's fathers and the role they played in the his- 
tory of the First Intermediate Period", ASAE 55 (1958), 167-190. 

13 W. Schenkel in: D. Arnold, Grdber des Alien und Mittleren Reiches in El-Tarif (Mainz: 
AV 17, 1976), 50. 

14 Schenkel, Memphis, 263-268 nos. 441-444. 

15 J.v. Beckerath, "Zur Begriindung der 12. Dynastie durch Ammenemes I.", ZAS 
92 (1965), 8-9. 


While none of the Manethonian sources lists individual kings, all of 
them claim that Dyn. 11 comprised 16 kings who ruled for 43 years. 
Evidently, the number of kings was inflated by 1 from 6 to 16, while 
the number of years was reduced by 100 from 143 to 43. Nevertheless 
the basic similarity of the figures again shows how close Manetho's 
data are to earlier pharaonic tradition. 

Unfortunately, the date of the re -unification of Egypt cannot be deter- 
mined precisely within the reign of Nebhepetre c Mentuhotpe II. 16 A 
stela of one Inyotef from Thebes is dated to year 1 4 of Mentuhotpe II, 
"the year of the rebellion of Thinis". 17 This event is in all probability 
to be seen in the context of the war between Thebans and Herakleo- 
politans, which was fought, at least in its earlier phases, in the Abydos- 
Asyut region. An inscription of an official who governed the Heliopolitan 
nome, dated to year 41 of Mentuhotpe II, provides proof that by then 
Mentuhotpe II controlled all of Egypt. 18 

Information which could clarify the sequence of historical events in 
the period between these two dates is lacking. The phases and modal- 
ities of Mentuhotpe IPs conquest of the Herakleopolitan kingdom, as 
well as the development of his political aspirations and his propaganda, 
remain unknown. Nevertheless it seems likely that both the actual course 
of political events and the ideological dimension of the war between 
Thebans and Herakleopolitans did have some influence on how the 
end of the Herakleopolitan dynasty was chronologically fixed in later 
annalistic tradition. Therefore one should be well aware that not only 
the date of the historical event of the re-unification of Egypt is lacking 
from our documentation but that the historical process itself in its sub- 
stance eludes us. 

It is even more difficult to use indirect criteria in an attempt to fix 
the date of the re-unification of Egypt. The two alterations of Mentuhotpe 
IPs royal protocol 19 may be linked to the stages of his rule over Egypt. 
In particular his latest Horus name zml-tl.wj "uniter of the two lands", 
attested for the first time in regnal year 39, 2() invited direct historical 

16 On this issue see Franke, "Chronologie I", 133, and Gestermann, Kontinuitat, 35-42, 
with further literature. 

17 Clere & Vandier, Textes, 19 § 23; Schenkel, Memphis, 227. 

18 Gestermann, Kontinuitat, 42 n. 5. 

19 Beckerath, Handbuch, 78-79. 

20 Inscriptions in Wadi Schatt el-Rigal, Schenkel, Memphis, 207-208 no. 318 and 
320. Beckerath's doubts that the date belongs to the rock inscription of Mentuhotpe II 
(Chronologie, 141 with n. 632) are in no way convincing. 


interpretation. However, as Gestermann rightly pointed out, this is not 
at all certain and in fact all his Horus names imply a claim of domi- 
nance over the whole of Egypt. 21 Rather more significant seems to be 
the development of the structure of the royal protocol. While in the 
first stage, Mentuhotpe, like his Theban FIP predecessors, used only a 
Horus name and the titles nswt-bjt and z?-R c with personal name, he 
introduced in the second stage both a and the throne name 
Nebhepetre c , thus claiming full royal status. 22 Whether this move reflected 
only his aspirations or his actual taking over of power remains unknown, 
however. In addition, from a strictly chronological point of view, this 
discussion is bound to remain more or less fruitless, since we do not 
know when the changes in the royal protocol occurred. Speculating 
that the earlier, rather than the later change of the protocol might be 
linked to the end of Herakleopolitan rule over northern Egypt would 
only intuitively make it more likely that Mentuhotpe's victory had 
occurred in the earlier part of his reign. Strictly speaking, however, an 
uncertainty of about 25 years in fixing the date of the re -unification 
of Egypt remains until new sources become available. 

The Herakleopolitan Dyns. 9 and 10 

Determining the identity and length of Herakleopolitan rule is much 
more difficult. 2 ' 5 Manetho's account listed two dynasties of rulers from 
Herakleopolis, 24 Dyn. 9 with four (Eusebius) or 19 kings (Africanus) 
who ruled for 100 (Eusebius) or 409 years (Africanus), and Dyn. 10 
with 19 kings who ruled for 185 years according to all sources. The 
TC, in contrast, listed only a single dynasty of 18 rulers; 25 unfortunately 
nearly all of the royal names and all reign length data as well as the 
total for this dynasty, which originally was given in line V. 10, are lost. 

21 Gestermann, Kontinuitat, 35-39. 

22 This view was envisaged already by Hayes and Arnold, see Gestermann, Kontinuitat, 
37, n. 2-3. 

23 See J.v. Beckerath, "The Date of the End of the Old Kingdom in Egypt", JNES 
21 (1962), 140-147; idem, "Die Dynastie der Herakleopoliten (9./10.)", ^4S 93 (1969), 
13-20; H. Goedicke, "Probleme der Herakleopolitenzeit", MDAIK 24 (1969), 136-143; 
Beckerath, Chronologie, 143-145; S. J. Seidlmayer, "Zwei Anmerkungen zur Dynastie 
der Herakleopoliten", GM 157 (1997), 81-90, with additional literature. 

24 Waddell, Manetho, 60-63. 

25 TC IV. 18-V. 10. 


The king lists of Abydos and Saqqara omitted the period of Herakleo- 
politan rule completely. 

While following the authority of the more ancient king list the unity 
of the Herakleopolitan dynasty was already assumed by Schenkel and 
others, 26 Malek was able to account convincingly for the discrepancy 
between the TC and Manetho. He demonstrated that the first four 
kings of the dynasty came to be separated from the main group of 
kings as a result of a series of misunderstandings which occurred when 
the original text was copied. 27 The number of 19 kings which Manetho 
gives for Dyn. 10 and which Africanus duplicated also for Dyn. 9 accu- 
rately reflects the total number of kings as given for the Herakleopolitan 
dynasty in TC V. 10. The difference of one king between TC and 
Manetho can very likely be explained by suggesting that the TC omitted 
the last Herakleopolitan ruler, just as the last king of Dyn. 1 1 was 
omitted, probably because he was not considered legitimate, having 
been removed from power under shameful circumstances as a result of 
the victory of the Thebans over the Herakleopolitan kingdom. 28 According 
to this analysis which interprets the separation of Dyns. 9 and 10 merely 
as a result of textual corruption in post NK tradition, the terms "Dyn. 
9" and "Dyn. 10" should, for the sake of terminological precision, no 
longer be used in historical interpretation to designate an earlier and 
a later phase of Herakleopolitan rule. 29 This argument has, of course, 
no implications whatsoever for the length and historical structure of 
the period. 

Since the relevant entries are destroyed in TC, and since only very 
few contemporary monuments of the Herakleopolitan kings are pre- 
served, the names and the sequence of the 19 Herakleopolitan kings 
cannot be reconstructed coherently. 30 Manetho names a king Khety as 
the founder of the dynasty, and the fact that the Herakleopolitan king- 
dom was referred to as pr Hty "the house of Khety" in contemporary 

2b Schenkel, Studkn, 149-150; Beckerath, "Herakleopoliten" (n. 23). 

27 J. Malek, "The Original Version of the Royal Canon of Turin", JEA 68 (1982), 
105; Redford, King-Lists, 238-239, arrives at an equivalent conclusion, though with 
different arguments. 

28 For another solution see Malek, "Version" (n. 27), 105. 

29 Seidlmayer, "Anmerkungen" (n. 23), 85-86. 

30 For the available data see Beckerath, Handbuch, 72—75; to the material listed there 
add M. Abd el-Gelil, A. Saadani & D. Raue, "Some Inscriptions and Reliefs from 
Matariya", MDAIK 52 (1996), 146-147. 


sources 31 lends credibility to this tradition. For the third ruler, the name 
Nfr-ki-K is preserved in TC. He could be identical with King Nfr-k>-R c 
whose name appears (with a graphic transposition kl-nfr-K, typical for 
the Herakleopolitan period) 32 on a label in the tomb of c Ankhtifi at 
Moalla. 33 The last ruler of the dynasty was probably Merykare c , whose 
name appears in texts relating to the final phase of the Theban- 
Herakleopolitan war in Siut tomb IV 34 and whose pyramid complex at 
Saqqara is attested in the titles of funerary priests. 35 In fact, the gen- 
eral tenor of the "Teachings for king Merykare"' seems to suggest that 
this king suffered defeat by the Thebans. 

Determining the length of the Herakleopolitan period is fraught with 
all but unsurmountable difficulties. Data for the length of individual 
reigns or of the entire dynasty are not preserved. Since it is clear, how- 
ever, that the Herakleopolitan dynasty did not start after Dyn. 1 1 but 
ran parallel to it up to the re-unification of Egypt under Nebhepetre c 
Mentuhotpe II, it must have lasted for between 87 and 114 years at 
least, depending on the exact date of the re-unification. Dismissing the 
data given by Manetho as unreliable, Beckerath developed the hypoth- 
esis of an ultra-short Herakleopolitan period making the beginning of 
Dyn. 1 1 more or less coeval with the establishment of the new Herak- 
leopolitan line of rulers in the north. 36 Apart from speculations about 
the historical development, this hypothesis rests mainly on the fact that 
so very few monuments are attested for the Herakleopolitan period, 
implying a short duration. This argument, however, is clearly not valid, 
since the 100 years or so when the Herakleopolitan dynasty existed 
parallel to the Theban kingdom did not leave many traces in the mon- 
umental record either. Evidently, this situation is due to the bias of 

31 E.g. the stela of Djari, Clere & Vandier, Textes, 14 § 18, 1. 3; see also O. D. Berlev, 
"The Eleventh Dynasty in the Dynastic History of Egypt", in: D. W. Young, ed., 
Studies presented to Hans Jakob Polotsky (East Gloucester/Mass., 1981), 361-377. 

32 Schenkel, Studien, 150; D. B. Spanel, "The Date of Ankhtifi of Mo'alla", CM 78 
(1984), 89. 

33 J. Vandier, Mo'alla (Cairo: BdE 18, 1950), 36. 

34 Schenkel, Memphis, 86 No. 64. 

35 PM (2) III, 562-563; J. Malek, "King Merykare and his Pyramid", in: C. Berger, 
G. Clerc & N. Grimal, eds., Hommages a Jean Leclant IV (Cairo: BdE 106.4, 1994), 
203—214; Kh. A. Daoud, "The Herakleopolitan Stelae from the Memphite Necropolis", 
in: Eyre, Proceedings, 303-308. 

36 Beckerath, "Herakleopoliten" (n. 23); for the details of the discussion see Seidlmayer, 
"Anmerkungen" (n. 23), 82-83. 


our sources, which tend to favour UE, and to the special character of 
FIP kingship which was no longer able to muster the resources of the 
country to construct monumental buildings. The argument that the 
Herakleopolitan kings were dropped from the Abydos king list and 
therefore never ruled UE does not carry much weight either. Even if 
the Herakleopolitan dynasty never wielded effective rule over UE, it could 
nevertheless have held nominal supremacy over the whole country for 
a period of time. The fact that a Herakleopolitan ruler was mentioned 
in the tomb of c Ankhtifi and that a certain Setka in his still unpub- 
lished biographical inscription in his tomb on Qubbet el-Hawa claimes 
to have entertained peaceful trade relations with the "House of Khety" 
seem to provide direct proof that there was indeed a period of 
Herakleopolitan sole rule before the advent of Dyn. 1 1 at Thebes. 

The only source which could provide data on the length of this 
period is the account of Manetho which has to be used with extreme 
caution, however. As was argued elsewhere, 3 ' Manetho's data for the 
length of Dyn. 9 can be dismissed since these numbers clearly could 
have been created only after the artificial division of the dynasties had 
occurred in post NK textual tradition. The information on the length 
of Dyn. 10, however, derives from a year total which was present 
already in earlier annalistic tradition and therefore could be of real his- 
torical value. Taking Manetho's figure at face value, the period of 
Herakleopolitan rule before the start of Dyn. 1 1 would amount to some- 
thing between 71 and 98 years, again depending on the date of the 
re-unification of the country. Of course, there is no way to be sure 
about the correctness of Manetho's figure; if one chooses to disregard 
Manetho's data, however, the length of the Herakleopolitan dynasty 
becomes entirely a matter of speculation, since there are no other 
sources available for fixing the length of Herakleopolitan rule before 
Dyn. 1 1 . Nevertheless a number of excellent scholars have tried, hypo- 
thetically linking the prosopographical and historical data which can 
be derived from tomb inscriptions from UE, to reach a reconstruction 
of the historical events of the earlier part of the FIP which would pro- 
vide the basis for a tentative estimate of its length. 315 It should be not 
overlooked, however, how very problematic this type of reasoning is 

37 For details see Seidlmayer, "Bemerkungen" (n. 23), 86-88. 

38 An intermediate estimate between Beckerath's minimum solution and the Mane- 
thonian maximum model was proposed e.g. by Fischer, Dendera, 131 n. 579. 


bound to remain and how very unlikely it is that the dispersed pieces 
of fragmentary evidence which were preserved by mere chance could 
indeed link up to form a historically coherent picture. 

The customary "short model" and the Manethonian "long model" 
result in two profoundly different views about the FIP. Therefore a 
look at the documentation for the period beyond the material relating 
to dynastic history in the strict sense could provide indications whether 
a short or a long model is more likely. Basing the argument solely on 
material from the royal sphere, as is often done in discussions of dynas- 
tic chronology, leads inevitably to a foreshortened perspective for peri- 
ods of weak kingship. The extensive prosopographic data from the FIP 
led Brovarski and Spanel to conclude that a succession of several gen- 
erations of local administrators held office in many UE towns between 
the end of the OK and the beginning of Dyn. 1 1, 39 thus clearly favour- 
ing a long model for the period in perfect accord with the data of 
Manetho. Also, the archaeological record can be made to bear on the 
problem. As was argued by Ward and Seidlmayer, the large number 
of burials in Upper Egyptian cemeteries which are to be dated to the 
earlier part of the FIP, as well as the fundamental morphological change 
which can be discerned in the archaeological material exactly in this 
phase, argue for a period of several generations. 40 Therefore, substan- 
tial evidence seems to support Manetho's figure for the length of the 
Herakleopolitan period. Nevertheless, the chronology remains on shaky 
ground. Since the discovery of new historical sources cannot be pre- 
dicted (although the excavations at the site of Herakleopolis Magna 
could certainly turn up new evidence), and since the potential of radio- 
carbon dating for this period seems to be limited, current efforts to 
establish a dendrochronological series which covers the second millen- 
nium BC might offer the prospect of substantial progress. 41 

39 E. Brovarski, "The Inscribed Material of the First Intermediate Period from Naga- 
ed-Der", AJA 89 (1985), 581-583; idem, The Inscribed Material of the First Intermediate 
Period from JVaga-ed-Der (Diss. Chicago, 1989) I, 15-53; D. B. Spanel, "The Date of 
Ankhtifi of Mo'alla", in GM 78 (1984), 87-94; idem, Bed Hasan in the Herakleopolitan 
Period (Diss. Toronto, 1984). 

40 W. A. Ward, Egypt and the East Mediterranean World 2200-1900 BC (Beirut, 1971), 
10—11; S. J. Seidlmayer, Graberfelder aus dem Ubergang vom Alten zum Mittleren Reich 
(Heidelberg: SAGA 1, 1990), 378; idem, "Anmerkungen" (n. 23), 84. 

41 See below Cichocki, chapter III. 3, with additional literature. 


Thomas Schneider 

1. Terminology and Methodology 

Substantial progress has been made in the study of the relative Chronology 
of the MK and Hyksos Period (MK/SIP) 1 since the presentation of the 
state of our knowledge fifteen years ago by D. Franke 2 thanks to new 
discoveries and the reappraisal of core issues. Moreover, there have 
been impulses to improve the terminological precision — particularly due 
to K. S. B. Ryholt's reassignment of dynasty designations — that are 
basically desirable. 

Ryholt 3 has reassigned the term "Dyn. 16", which had traditionally 
been used to designate vassals of the Dyn. 15 Hyksos, although it was 
known that this was based on an erroneous reading of the Manethonian 
tradition, 4 to a sequence of Theban rulers which would have been listed 
in TC X, 31— XI, 14 and which he identifies as a "First Theban Dynasty".'' 
While the term "Dyn. 17" is generally understood as referring to all 
of the Theban rulers between Dyns. 13 and 18 who were believed to 
be listed in TC X, 31— XI, 14, this term is used by Ryholt only for a 

1 The designation of the period is a historiographical, rather than a chronological, 
concern. Cf. Franke, "Chronologie II", 245-274, esp. 245f.; idem, Heqaib, 77-78; 
Schneider, Auslander, 155-156; differentiy, Ryholt, Situation, 311. For the present chap- 
ter, I prefer the tide "MK and Hyksos Period", as Ryholt's use of "Intermediate Period" 
is based on the alleged political division of Egypt since the end of Dyn. 12, which is 
not necessarily valid (cf. below) whereas our term corresponds to the period. I would 
like to thank M. Bietak, D. Franke, E. Hornung, D. Polz and A. Spalinger who read 
and commented upon earlier drafts of the text. Bibliography has been considered until 
spring 2002 when the manuscript was delivered. 

2 Franke, "Chronologie I. II." 

3 Ryholt, Situation. Reviews: D. Ben-Tor, S. Allen, J. P. Allen, "Seals and Kings", 
BASOR 315 (1999), 47-74; R. Holton Pierce, Acta Orientalia 60 (1999), 207-213; J. v. 
Beckerath, AfO 46/47 (1999/2000), 433-435; W. Grajetzki, OLZ 95(2000), 149-156; 
A. Dodson, BiOr 57(2000), 48-52; A.S. Spalinger, JNES 60 (2001), 296-300. 

4 Beckerath, Untersuchungen, 17-20, Schneider, Auslander, 123. 

5 Ryholt, Situation, 151. 


chronologically later group of Theban kings, his "The Second Theban 
Dynasty". The rulers of Near Eastern origin, but not members of Dyn. 
15, are then designated as "Dyn. 14". b 

In a response, Beckerath has suggested that the entire sequence of 
Theban kings between Dyn. 13 and 18 should henceforth be termed 
"Dyn. 16", and to use the term "Dyn. 17" for the rival dynasty at 
Abydos postulated by Ryholt. 7 Depending upon the course and out- 
come of the discussion (for details, cf. below section 4) it might be use- 
ful to maintain "Dyn. 17" for the (larger number of) Theban kings. 

In order to avoid the menace of conceptual confusion, this contri- 
bution will distinguish the different terminologies by adding the expo- 
nents " T " (for traditional) and " R " (for Ryholt) to positions of kings where 
these vary in the chronological reconstruction. Furthermore, alternative 
dynasties differing from the customary usage will be put in quotation 
marks, e.g. Dyn. 16 signifies the conventional Dyn. 16, whereas "Dyn. 
16" refers to the terminological reassignment by Ryholt or later. The 
numbering of the columns in the TC follows Gardiner's standard pub- 
lications whereas different systems of numbering are explicitly stated 
("Ryholt's ninth column"). 

Methodologically, it is important to note that a chronological frame- 
work for the period can be established that extends from Dyn. 12 to 
Dyn. 13 where it can be roughly fitted into Dyns. 15 and 17 (or "Dyns. 
16/17" according to Ryholt). In its original state, the TC recorded at 
least 50 kings for Dyn. 14, for whom, however, the length of the reign 
is preserved for only a few, and only two ("-sh-r c Nhsj; Mrj-df'-r 3 ) are 
otherwise attested through monuments. The rulers known from scarabs 
are not among those listed in the entries for Dyn. 14 in the TC. Ryholt's 
attempt to create a typological sequence of scarabs upon which to 
build chronologically acceptable successions has encountered consider- 
able criticism." As other hypotheses are likewise difficult to support (an 
overlap of early Dyn. 14 and the end of Dyn. 12; 9 the relationship be- 
tween Dyns. 13 and 14 in the sense of a "trade agreement"), 10 Dyn. 14 

6 Ryholt, Situation, 94ff. 

' Cf. below, section 4. A decision depends largely upon the interpretation of TCXI. 

8 Ben-Tor et al. (n. 3), 53-65. 

9 Ben-Tor et al. (n. 3), 55, 59, 66. 

10 Ben-Tor et al. (n. 3), 59f; Spalinger (n. 3), 299; more positive, Beckerath (n. 3), 
434; Grajetzki (n. 3), 153-154 (who does point out that of 600 seals from Dyn. 14, 
only one was found in Ryholt's postulated capital of Dyn. 14, Avaris/Tell el-Daba). 


disappears as an alternative to Dyn. 13 as a chronological link between 
the MK and Dyns. 15 and 17 ("Dyns. 16/17" according to Ryholt). 
Ryholt weeds out a number of fragments of TC col. X which allegedly 
record fictitious royal names, and places them in a postulated addi- 
tional column of gods and demigods at the start of the papyrus (new 
second column) with the result that the numbering of the columns 
would increase by one for all following columns. 11 

2. Dynastie 12 

The more recent chronological discussion on Dyn. 12 has been dom- 
inated by the issue of coregencies, whereas the succession of rulers 
and their reign -lengths are in principle resolved. 12 After Franke in his 
survey in 1988 assumed the validity of the coregencies of Dyn. 12, 1!i 
C. Obsomer has not only rejected the alleged 10-year coregency of 
Amenemhet I and Senwosret I, but all other coregencies of Dyn. 12 
as well, in the extensive discussion of the problem in his study of the 
reign of Senwosret I (as R. D. Delia and W. Helck before him). 14 Delia 
remarked about this: "Obsomer's reconstruction largely rests upon an 
all-or-nothing foundation. If one is unconvinced that Obsomer has sat- 
isfactorily explained away all of the coregency evidence, then much of 

11 Ryholt, Situation, 24-25. In some responses to Ryholt's work, the elimination of 
the allegedly fictitious names of the TC have been met with enthusiastic agreement: 
Ben-Tor et al. (n. 3), 315, 49; Beckerath (n. 3), 433; idem, Handbuch, 282; Dodson 
(n. 3), 49. Note however, that Ryholt's proposal (Situation, 24 n. 59) takes part of its 
legitimation from considering Redford's efforts to recognize West Semitic names in the 
notations a failure. A modified version of the approach has been offered by Schneider, 
Auslander, 99-122; the doubts mentioned there (100) on the reorganization of the frag- 
ments have not been eliminated. There is also the question whether names such as 
"Protector of noble women" etc. would be more plausible among the mythical demigods. 
Note also that Ryholt moves fr. 42 with the divine name "Apis" whereas fr. 123 with 
the same name "Apis", followed by two entries written in group writing, is left in the 
section of the SIP. Fr. 22 which Ryholt also shifts might preserve the name of a ruler 
known from scarabs, 'nt-hr (Schneider, Auslander, 134), and fr. 152 another king known 
from scarabs, Pns (Schneider, Auslander, 106-109, 140). 

12 An absolute chronological date in the early 12th dynasty could perhaps be pro- 
vided by the astronomical reference of Sesostris's temple at Karnak, cf. L. Gabolde, 
Le „Grand Chateau d'Atnon de Sesostris Ier a Karnak" (Paris, 1998), 123-134; but see below 
Belmonte, Chapter III. 5. 

13 D. Franke, Chronologie II", 114-125. In idem, Heqaib, XII, he shares the view 
rejecting the coregency of Amenemhet I and Senwosret I. 

14 C. Obsomer, Sesostris I". Etude chronologique et historique du r'egne (Bruxelles, 1995). 


the structure of his history and chronology crumbles although many 
gems remain in the details." 13 After a detailed investigation of Obsomer's 
arguments and the entire situation, K. Jansen-Winkeln came down clearly 
in favour of coregencies in 1997. 16 

In fact since Franke's summary, a number of new documents and 
archaeological evidence, particularly from the sites of the pyramid com- 
plexes of Dyn. 12, has appeared which objectively speaking can hardly 
be interpreted in a different manner. In the fundamental case, of the 
first royal transition from Amenemhet I to Senwosret I, the 10-year 
coregency of these first two kings is favoured not merely by the two 
main witnesses to which appeal is usually made (Stela Cairo CG 20516 
naming both rulers in the field at the top with the regnal years "30" 
and "10" and Louvre CI with a date naming both kings). 17 An archi- 
trave from Matariya published in 1990 names both kings symmetri- 
cally with their titularies and apparently as co-reigning builders; both 
are designated as nsw bjt and living Horus (i.e. as reigning king). 1 " Finally, 
the control marks from Lisht published by F. Arnold reveal that it was 
only in regnal year 10 of Senwosret I that the construction of his pyra- 
mid began, i.e., apparently after the death and burial of Amenemhet 
I in his pyramid complex. 19 The description of the attack on Amenemhet 
I in the Teaching of Amenemhat I may have served, as Jansen-Winkeln 
suspected, 20 to legitimize the introduction of the institution; making, in 
any case, "a decision in favour of a coregency practically unavoid- 
able". 21 The length of the reign of Senwosret I is somewhat more than 

15 R. D. Delia, JARCE 34 (1997), 267-268; sceptical is also C. Gallorini, BE 39 
(1997), 135-137. 

16 K. Jansen-Winkeln, "Zu den Koregenzen der 12. Dynastie", SAK 24 (1997), 
115—135. Beckerath, Chronologie, 133, also comes out in favour of the coregencies. 

17 Jansen-Winkeln (n. 16), 122-125. 

18 Both are also wished "eternal life", which clearly underscores that Amenemhet I 
could not have been dead at the time (cf. Ryholt, Situation, 273); A. Awadalla, "Un 
document prouvant la coregence d'Amenemhat et de Sesostris I", CM 115 (1990), 
7-14; Jansen-Winkeln (n. 16), 125. 

19 F. Arnold, The Control Motes and Team Marks (New York: PMMA 23, 1991) 19ff, 
30ff.; Jansen-Winkeln (n. 16), 125-126. 

20 K. Jansen-Winkeln, "Das Attentat auf Amenemhat I. und die erste agyptische 
Koregentschaft", SAK 18 (1991), 241-264; idem (n. 16), 128-135. Cf., differentiy, 
N. Grimal, "Coregence et association au trone: l'Enseignement d'Amenemhet I cr ", 
BIFAO 94 (1994), 143-172. 

21 L. M. Berman, Amenemhet I {New Haven: Ph.D. Yale, 1985), 173-213. R. Leprohon, 
"The Programmatic Use of the Royal Titulary in the Twelfth Dynasty", JARCE 33 
(1996), 167 now supports the coregency. 


45 years (as deduced from the highest attested date in a graffito south 
of Amada and the entry in the TC, whereas Manetho has 46 years). 
A coregency of roughly 3 years with his successor can be plausibly 
deduced from the stela Leiden V.4, where apparently, regnal year 44 
of Senwosret I is made equivalent to regnal year 2 of his son Amenemhet 
II. 22 A still more explicit equivalency of regnal years using the prepo- 
sition hft comes from the coregency of Amenemhet II and Senwosret 
II, whose regnal year 3 is identified as the regnal year 35 of his pre- 
decessor in a stela near Konosso. 23 This date is also the highest recorded 
regnal year for Amenemhet II. 

A regnal year 8 (9?) is preserved for Senwosret II on a stela from 
Toshka; and a "year 8" is also now known from the stela Cairo JE 
59485. 24 It is therefore necessary to emend accordingly the Manethonian 
tradition and the figure of "19 years" in TC VI, 23 (see below). There 
would not appear to have been a coregency with his successor Sen- 
wosret III. 

The debate over a short (19 year) or a long (39 year) reign for 
Senwosret III can now be viewed as resolved. A regnal year "39" found 
in the valley temple of the cenotaph of Senwosret III at Abydos in 
1994 confirms a long reign, 25 supported by a control mark of year "30" 
from the royal pyramid complex at Dahshur, 26 and a reference to a 
first ^-festival of this king 2 ' — as had already been argued by W. Helck, 
W. Barta and J. v. Beckerath; TC VI, 24 indicates 30+ regnal years. 
In terms of relative chronology, this is not relevant as there had been 
a coregency with Amenemhet III since year 20, 28 and thus regnal year 
20 of Senwosret III is identical to regnal year 1 of his son. It is in this 

22 Jansen-Winkeln (n. 16), 117-118. 

23 Jansen-Winkeln (n. 16), 118—119. See also W. K. Simpson, "Studies in the Twelfth 
Egyptian Dynasty III: Year 25 in the Era of the Oryx Nome and the Famine Years 
in Early Dynasty 12", JARCE 38 (2001), 7f: Regnal year 43 of Senwosret I corre- 
sponds to (hft) year 25 of the nomarch Amenemhet. 

24 M. C. Stone, "Reading the Highest Attested Regnal Year Date for Senwosret II: 
Stela Cairo JE 59485", GM 159 (1997), 91-99. 

25 J. M. Wegner, "The Nature and Chronology of the Senwosret III — Amenemhet 
III Regnal Succession: Some Considerations Based on New Evidence from the Mortuary 
Temple of Senwosret III at Abydos", JJVES 55 (1996), 249-279; cf idem, The Mortuary 
Complex of Senwosret II (Philadelphia: PhD Thesis, 1996), 416. 

26 F. Arnold, "New Evidence for the Length of Reign of Senwosret III?", GM 129 
(1992), 27-31. 

2/ D. Arnold & A. Oppenheim, "Reexcavating the Pyramid Complex of Senwosret 
III at Dahshur", KMT 6/2 (1995). 
28 Jansen-Winkeln (n. 16), 119-120. 


fashion that one must understand the change of dates in pBerlin 10055 
from Ulahun where a "year 19" is followed by a "year 1", which has 
previously been used by those favouring a short reign (M. Bietak, 
D. Franke) and would entail an emendation in Manetho and the TCP 
Chronologically, it is these 19 years that must be accounted for in a 
relative chronology. 

Until recently, the alternatives for Amenemhet III did not affect 
chronology: whether proposing a short reign for Senwosret III without 
a coregency (Franke) or a long reign with a 20-year coregency. With 
Senwosret's long reign confirmed, a chronological consequence would 
result if one rejected the coregency (as do Delia 50 and Obsomer). 51 
However, the coregency is also confirmed by the presence of the names 
of both kings on scarabs and cylinder seals as well as fragments of 
inscriptions of the coronation ritual of Amenemhet III which was appar- 
ently performed by Senwosret III. 32 

The highest explicit regnal year attested for Amenemhet III is the 
45th, but a "year 46" in the Ulahun letters should probably also be 
assigned to him. A short coregency with Amenemhet IV is assured by 
a double date in a rock inscription at Semna (RIS 7), where the first 
regnal year of Amenemhet IV is made equivalent to regnal year 44 
(or 46 or 48?) of Amenemhet III. The coregency is also supported by 
representations of two kings from the pyramid complex of Amenemhet 
III in Hawara. 33 

The TC records a reign of 9 years, 3 months and 27 days for 
Amenemhet IV. The highest date known from inscriptions is that of 
the 9th year (Sinai 122), but a "year 10 (?)" from the Ulahun papyri 
may belong to him as well. His successor Nofrusobek reigned for 3 years, 
10 months and 24 days according to the TC. M Her highest date doc- 
umented epigraphically is regnal year 3 on a Nile level mark at Kumma. 

29 See also Ryholt, Situation, 212 n. 728, who still assumes a short reign for Senwosret 
III and a one-year coregency with Amenemhet III. 

30 R. D. Delia, A Study of the Reign of Senwosret III (New York: PhD Thesis, 1980), 

31 Obsomer (n. 14). 

32 Jansen-Winkeln (n. 16), 120; Ryholt, Situation, 212 n. 728. 

33 Cf. Franke, "Chronologie," 120; Ryholt, Situation, 209-210 with n. 716. This core- 
gency is also supported by Leprohon (n. 21), 170. 

34 For her, cf. also V. G. Callender, "Materials for the Reign of Sebekneferu", in: 
Eyre, Proceedings, 227-236; S. Roth, Die Konigsmutter des Alten Agypten von der Fruhzeit bis 
zum Ende der 12. Dynastie (Wiesbaden: AAT 46, 2001), 242-245 (against Ryholt's hypoth- 
esis of a kinship and its implications). 



The original text of the TC seems to have corresponded to the dates 
known from epigraphical sources in almost every instance. The reigns 
preserved in the TC are: 

Amenemhet I 
Senwosret I 
Amenemhet IV 

(2)9 years 

(xj months 








Rests of the entries (without names) of the four reigns between Senswosret 
I and Amenemhet IV are probably preserved on fr. 67 and have been 
discussed most recentiy by Krauss, 35 Franke, 36 and Ryholt. 37 The four 
lines give these numbers of regnal years: 




10 (or 20/30?) + x 

30 (+ x) 
40 (+ x) 

Of these, only TC fr. 67,2 has to be emended (19 instead of 9). The 
entire picture would thus be: 




Coregency with Successor 

Turin Canon 

Amenemhet I 


10 years 

29 years 

Senwosret I 


2-3 years 

45 years 

Amenemhet II 


3 years 

10/20/30 + x years 

Senwosret II 



19 years 

Senwosret III 


20 years 

30 + x years 

Amenemhet III 


1 year (?) 

40 + x years 

Amenemhet IV 


9 y., 3 m., 

27 d. 



3 y., 10 m. 

, 24 d. 

Krauss, Sothis, 194-195. 
Franke, "Chronologie I," 122, 
Ryholt, Situation, 14-15. 



The total length of Dyn. 1 2 would be the sum of the reigns minus the 
coregencies, which amounts to ca. 181 years according to this table. 
This would confirm the year sum recorded by Eusebius of "182". By 
contrast, the sum in TC VI, 3 of "213 years, 1 month; 15/17/19 days" 
was simply the total of the individual reigns without subtracting the 
coregencies. 38 

3. Dynastie 13 

Ryholt's monograph provides a comprehensive new discussion of Dyn. 
1 3 which necessarily entails numerous shifts and reassignments of chrono- 
logical positions. TC VII, 5 — VIII, 27 listed a total of 51 kings, which 
he raises to 57, based on the restoration of lacunae which he postu- 
lates for the three wsf-notes in VII, 6, VII, 7 and VII, 17. 39 Manetho 
gives a total of 60 kings for Dyn. 13. In order to make the changes 
clear, I provide a table with both the traditional sequence (according 
to Franke) and the new one proposed by Ryholt in Fig. II. 7.1. 

According to TC VII, 5, a king Wegaf founded Dyn. 13. As earlier 
H. Stock and K. A. Kitchen, Ryholt shifts this king 20 places so that 
he follows Amenemhet VII, since the names of Wegaf appear on both 
sides of the titulary of Amenemhet VII on the back of a statue base 
dedicated to Month in Medamud. 40 However, the fact that the name 
Wegaf was later added to the statue does not necessarily demand the 
assumption of a later reign (after Amenemhet VII). As only about 30 
years separate the reign of Amenemhet VII and the foundation of the 
dynasty, it is also conceivable that a later reference to the founder of 
the dynasty by a king who may have been related to him was desired. 
In Ryholt's reconstruction, Sekhemre'-khutawy Sobekhotep I who is 
traditionally identified as Sobekhotep II and listed as the 16th king of 
the dynasty, opens Dyn. 13 which is justified by assuming an erroneous 
exchange of names in the TC A definite judgement on the matter is 
not possible at present. 41 

38 Ryholt, Situation, 16 (partially based on other figures; differently also Franke, 
"Chronologie I," 122, 126f.; Krauss, Sothis, 198; Beckerath, Chronologie, 134. 

39 Ryholt, Situation, 72. 

40 Ryholt, Situation, 317-318. 

41 Positive is Allen in: Ben-Tor et al. (n. 3), 50; c£, however, Franke, "Chronologie II", 



Franke, "Chronologie" II (1988) 

1 ) Khutawyre / Wegaf 

2) Sekhcmkare / Amenemhetsonbef 
= Amcncmhct V ? 

3) (Sekhemrekhutawy) 

4) * Amcncmhct (V) 

5) Sehetepibre 

6) Iuefni 

7) Seankhibre / Amenemhet VT 

8) Semenkare / Nebnun 

9) (Se-)Hetepibre / Harnedjheritef 
TO) Scwadjkare 

1 1 ) Ncdjcraibrc 

12) Khaianchre / Sobekhotep I 

13) Rcniscncb 

15) Sedjefakare / Amenemhet VII 

16) Sekhemrekhutawy / Sobekh. II 

17) Userkare /Khendjer 

18) Semenkhkare / Mermesha 

19) Schctcpkarc / Inyotcf IV 

20) ...ibrc/ Sethi 

21) Sekhemresewadjtawy / Sebekh. Ill 

22) Khasekhemre / Neferhotep I 

23) Menwadjre / Sihathor 

24) Khancfcn-c / Scbckhotcp IV 

25) Khahetepre / Sebekhotep V 

26) Wahibre / Jauib (Ibiau) 

27) Merneferre /Aya 

28) Mcrhctcprc / Ini (Scbckh. VI ?) 

Dvn. 13 

Unplaced: Qemau 
Unplaced: Scbckaj 

Ryholt, Situation (1997) 

1) Sobekhotep I Sekhemrekhutawy 

2) Senbef Sekhemkare 

3)- Nerikare 

4) Amcncmhct V Sekhcmkare 

5) Qemau - 

6) Siharnedjheritef Hetepibre 

7) Iuefni - 

R) Amenemhet VI Seankhibre 

9) Nebnun Semenkare 

10) Seusekhtaui (II) Sehetepibre 

11) - Scwadjkare 

12) - Nedjemibre 

13) Sobekhotep II Khaankhrc 

14) Reniseneb - 

16) Khabau (H) Sekhemrekhutawy 

1 7) Djedkheperu (H) ...kare 

19) Kay - 
\20) Amcncmhct VII Sedjefakare 

2 1 ) Wegaf Khutawyre 

22) Khendjer Userkare 

23) Mermesha Semenkhkare 

24) Inyotcf V Schctcpkarc 

25) Seth Meribre 

26) Sobekh. Ill Sekhemresewadjt. 

27) Neferhotep I Khasekhemre 

28) Sihathor Menwadjre 

29) Sobekhotep IV Khaneferre 

30) Sobekhotep V Mcrhotcprc 

31) Sobekhotep VI Khahotepre 

32) Ibiau Wahibre 

33) Aya Merneferre 

34) Ini Merholepre 

Fig. II. 7.1 


Ryholt postulates a crisis of legitimacy at the start of Dyn. 13, on 
the assumption that Dyn. 14 had recently come to power. The dou- 
ble names of early Dyn. 1 3 would indicate a filiation that was to legit- 
imize their rule. In this vein, a name such as Ameni Qemau would 
mean "Ameni's (Amenemhet's) son Qemau". However, the proposed 
early date for Dyn. 14 rests on very uncertain foundations and has 
probably to be rejected. 42 By contrast, the interpretation of the double 
names as filiations (without any further political implications) appears 
very plausible. 43 In the 2nd and 4th positions of the TC are "Amenem- 
hetsenbef" and "Amenemhet" (V). Whereas Ryholt proposes reading 
"Amenemhet (IV)'s son Senbef", earlier scholars had considered that 
the two were identical. 44 The assumption of a name indicating filiation 
has implications for other relative positions at the start of Dyn. 13. 
King Ameni Qemau, who had not been assigned a firm place until 
now, probably also occupied a position in early Dyn. 13. If his dou- 
ble name is to be understood as a filiation, Qemau could, following 
Ryholt, be understood as the son of Ameni = Amenemhet V, and 
assigned the 5th position. A more convincing solution is that proposed 
by J. P. Allen where the fragmentary entry TC VI, 7 "Amenemhet" is 
believed not to denote Amenemhet V, but Amenemhet Qemaw, and 
thus the lacuna of TC VI, 6 would once have held the name of 
Amenemhet V. 4;> The 5th position (Sehetepibre c ) is shifted to place 10 
by Ryholt. 

The 3rd position in Dyn. 13, where only a figure of 6 regnal years 
is preserved in the TC, remains unclear. Ryholt fills this with a Nerikare, 
known from a Theban stela, on the basis of a Nile level mark at Semna, 
but the Semna record should be read in a different fashion. 46 Beckerath 
proposes the hypothetical Pentjini here, who is a ruler in the parallel 
provincial line in Abydos according to Ryholt. 4 ' 

The following five rulers (places 6—9) are identical in both the tra- 
ditional arrangement and in the new reconstruction, with only the last 
two Har-nedjheritef and Sehetepibre c having been switched. The TC 

42 Cf. the references above in note 10. 

43 Beckerath (n. 3), 434; Allen in: Ben-Tor et al. (n. 3), 50; Dodson (n. 3), 50. 

44 Ryholt, Situation, 208, 212. 

45 Allen in: Ben-Tor et al. (n. 3), 50. 

46 Beckerath (n. 3), 434. The existence of this king had earlier been thrown into 
doubt by L. Gabolde, "Nerkare, a-t-il existe?", BIFAO 90 (1990), 213-222. 

4/ Beckerath, Chronologie, 137. 


records the throne name of Sehetepibre c for both so that an error must 
be in place in the case of Har-nedjheritef (with a throne name Hetepibre c ). 
The reason for Ryholt's proposed switch of Har-nedjheritef (read as 
Sihar-nedjheritef by him) is the assumption of a filiation, as Kmlw z) 
precedes the name in the cartouche. Ryholt takes the word "son" as 
the first part of the personal name, and interprets the sequence of the 
two names as a filiation: Qemaw's son Sihar-nedjheritef. Accordingly, 
he places the king immediately after his hypothetical father Qemaw. 
However, as the inverted filiation can be written with or without the 
word for "son", the proper name can still be read Har-nedjheritef; even 
as the son of Qemaw he could have come to power a few years after 
his father's reign. 4 " 

The next five positions are once again identical (Franke 10—14 = 
Ryholt 11—15). One should merely note that Sobekhotep bears the ordi- 
nal number "II" in Ryholt's reconstruction because the traditional 
Sobekhotep II has been shifted from the 16th place to the first. The 
last of these five kings is Hor (I) Awibre c , whose Horus name is attested 
on a reused block in Tanis together with the Horus name Khabaw. 
Ryholt therefore concludes that Khabaw Sekhemre c -khutawy was the 
successor of Hor (I) Awibre c (places 15 and 16), whereas the lost 3rd 
position of Dyn. 13 in the TC was customarily reserved for him. 49 

Accounting for the fact that neither does Khabaw Sekhemre'-khutawy 
follow Awibre c in the TC, nor does a certain Djedkheperu (Horus name) 
attested in seal impressions from Uronarti alongside Khabaw's, but 
immediately Amenemhet VII Sedjefakare c (TC VII, 18), Ryholt postu- 
lates a lacuna of four places in the Vorlage of the TC, to be filled by 
Khabaw, Djedkheperu, Seb and Kay. 50 To Djedkheperu he assigns the 
Osiris bed from Abydos,' 1 and suggests that the effaced titulary con- 
sisted in the name of Hor (I Awibre) as the beginning of his proper 
name and simultaneously filiation, but J. P. Allen has plausibly argued 
that it was Pentjini who was responsible for the inscription.' 2 

48 N. Swelim & A. Dodson, "On the Pyramid of Ameny-Qemau and Its Canopic 
Equipment", MDAIK 54 (1998), 330, suggest that Har-nedjheritef was only the fourth 
successor of his father. 

49 Critically, and with an alternative Ben-Tor et al. (n. 3), 50. 

50 Ryholt interprets the note wsfoi the TC as an indication of a lacuna in the Vorlage, 
not as a reference to a king whose name had been deliberately omitted. For this ques- 
tion, cf. C. Bennett, "King Qemau: a Reconsideration", CM 159 (1997), 11-17. 

51 This was assigned to Khendjer by A. Leahy and to Nebiriraw II by Beckerath. 
: ' 2 Allen in: Ben-Tor et al. (n. 3), 50-51. 


Ryholt postulates Set and Kay as new kings of Dyn. 13, as he dissolves 
the proper name Sebkay (Sbkly) on the magic ivory CG 9433 (JdE 
34988) into the filiation, "Seb's son Kay". Since Amenemhet VII names 
himself Kay-Amenemhet, three generations of kings would thus have 
reigned within the span of a few years. This "daring construction" 
(J. von Beckerath)'* which in addition requires the postulation of a 
lacuna in the Vorlage of the TC does not seem necessary. If we do not 
move Sobekhotep II from his traditional place 16 to the top of the 
dynasty (against Ryholt), the proper name of the magic ivory can be 
assigned to him. In this case we would have a pet form of a divine name 
formed by means of the suffix -li. In the NK this suffix was written 
<nr>, <nij< (Hali for Hathor, Wurel for Weret), 34 for which in Dyn. 13 
a notation with <>> would have been mandatory. The sounding of the 
name would thus have been something like Sabkuli. The possible 
patronymic for Amenemhet (K>ji — Kuli) would be the abbreviated form, 
so that Sobekhotep II (and not an otherwise unknown K>y) should be 
viewed as the father of Amenemhet. This would require a switch between 
TC VII, 18 and TC VII, 19 at the most. Amenemhet VII is followed 
in the traditional chronology by Sobekhotep (II) Sekhemre-khutawy, 
and by Wegaf who has changed places with him according to Ryholt. 

This takes us into uncontested territory: the following 16 kings of 
Dyn. 13 are arranged in the same succession in the two competing 
reconstructions, with a single minor variation: Sobekhotep Merhotepre c 
would be inserted as Sobekhotep V between Sobekhotep IV Khaneferre c 
and Sobekhotep VI (previously labelled 'Sobekhotep V), and his 
identification with Ini Merhotepre c completely abandoned. 35 A relief 
from the reign of Neferhotep I depicting the enthroned prince of Byblos 
Antin serves as the traditional basis for the synchronism between Egypt 
and Babylonia for the first half of the second millennium BC (Neferhotep: 
Jantin'ammu of Byblos: Zimrilim of Mari: Hammurabi), 56 which is, 
however, probably not reliable. 5 ' 

53 Beckerath (n. 3), 434; a similar appreciation by Grajetzki (n. 3), 155; cf. Allen in: 
Ben-Tor et al. (n. 3), 51. 

54 T. Schneider, Asiatische Personennamen in agyptischen Quellen des Neuen Reiches (Freiburg: 
OBO 114, 1992), 276-277. 

55 Ryholt, Situation, 231-232. 

56 E. Hornung, "Lang oder kurz? — Das Mittlere und Neue Reich Agyptens als 
Priifstein", in: High, Middle or Low? 1, 27-36; K. A. Kitchen, "The Basics of Egyptian 
Chronology in Relation to the Bronze Age", in: High, Middle or Low? 1, 37-55; Franke, 
"Chronologie II", 273-274; Ryholt, Situation, 87-88. 

'" C. Eder, Die dgyptischen Motive in der Glyptik des ostlichen Mittelmeerraumes zu Anfang des 


Aya Merneferre c is the last ruler of the dynasty attested in both the 
north and the south, whereas his successor Ini Merhotepre c is the first 
to be only attested in southern Egypt. The lower part of TC's column 
VII (Ryholt's eighth column) — after Sobekhotep VII Merikawre c ; TC 
VII, 8 — is very badly preserved after a long gap. At the end of the 
column (now lost) was probably the total for Dyn. 13 and the start of 
Dyn. 14, which continued on column VIII (Ryholt's ninth column), 
with Nehesi in TC VIII, l. 58 That Nehesi should have had five prede- 
cessors which would justify the postulation of an additional lacuna in 
the Vorlage of the 7"C 59 is hardly plausible. 60 

Of the lengths of the reigns of Dyn. 13 kings in the TC, the fol- 
lowing entries display preserved year dates: 61 

TC VI, 5 


2 years, 3 months, 24 days 

TC VI, 6 


6 years 

TC VI, 7 

Amenemhet V 

3 or 4 years 

TC VI, 8 


1—4 years 

TCVI, 14 


years, 7 months 

rcvi, 16 


years, 4 months 

TC VI, 24 

Sobekhotep III 

4 years, 2 months 

TC VI, 25 

Neferhotep I 

11 years, 1-4 months 

TC VI, 26 


years, 1 + x months, 3 days 

rc* vii, i 

Sobekhotep VVVP 

4 years, 8 months, 29 days 

TC VII, 2 


10 years, 8 months, 28 days 

TC VII, 3 


23 years, 8 months, 18 days 

TC VII, 4 


2 years, 2-4 months, 62 9 days 

TC VII, 5 


3 years, 2-4 months 

TC VII, 6 


3 years, 1 months, 1 day 

TC VII, 7 


5 years, ? months, 8 days 

TC VII, 8 

Sobekhotep VII 

2 years, ? months, 4 63 days 

2 Jts. v.Chr. (Leuven: OLA 71, 1996), 13; T. Schneider, £DPF 114 (1998), 184-188. 
Jantin itself would have been a hypocoristic abbreviation of Jantin'ammu, but what 
we have is only 'An tin. 

58 For Nehesi, see now A. Loprieno, "Misj, 'der Siidlander'?", in: Stationen, 211—217. 

59 Ryholt, Situation, 94. 

60 See above, notes 8-10. Spalinger (n. 3), 297 assumes (following Ryholt) that it is 
demonstrated that Sheshi belongs in early Dyn. 14, whereas this assignment is dis- 
puted by Ben-Tor et al. (n. 3), 61. 

bl Cf Ryholt, Situation, 192, and Fig. 10 with the sheet joins as opposed to Franke, 
"Chronologie IF', 267ff. 

1,2 Franke, "Chronologie II", 267ff. has "2" months, Ryholt "3-4"; cf. however TC 
VIII, 5, and thus "2-4". 

b:i Ryholt, Situation, 192 has erroneously "3" days. 


Down to the reign of Aya, a minimum of 1 00 years must be accounted 
for since the beginning of Dyn. 13. The last ca. 25 kings of Dyn. 13 
are very badly documented in the epigraphic material, and their mon- 
uments are restricted to a few places (Abydos, Thebes, Deir el-Bahri, 
Gebelein, Edfu). "By the end of Aya's 24-year reign, the administra- 
tion seems to have collapsed [. . .] This period further witnessed no less 
than 24 kings, whose reigns on the average lasted little more than one 
year.'"' 4 In place of a precise chronology, we can merely estimate a 
total of some two decades, after which Dyn. 1 3 would be immediately 
followed by the new Theban dynasty (Dyn. 17 according to the tra- 
ditional version; 6 * 1 "Dyn. 16" according to Ryholt). 66 There are some 
dated monuments for a few kings data for whose reigns are not pre- 
served in the TC: Sobekhotep II T /I R (year 4), Amenemhet-Senbef (year 
5), Khendjer (year 5), Sobekhotep IV (year 9) and a few additional of 
contested attribution. 67 The only indication of the total length of Dyn. 
13 remains the total given by Manetho of 453 years, which is usually 
emended to *153 or 133 (Beckerath) years and represents the basis for 
approximate estimates of 133 years (Beckerath), 68 some 130 years 
(Franke), 69 152 years (Kitchen), 70 and 154 years (Ryholt);' 1 cf. our final 
remarks in section 6. 

4. Dyn. 17 (Ryholt: "Dyn. 16" & u Dyn. 17") 

4. 1 . Generalities Arranging the relative chronology of the kings attested 
in the Thebaid between Dyns. 13 and 18 is one of the most difficult 
challenges of the SIP. The traditional reconstruction of Dyn. 1 7 assumes 
15 rulers, based on the sum in TC XI, 15 (which must apparently be 
emended to "15"); these kings would have been named in TCX, 31— XI, 
14. 72 Against this, Ryholt (following H. E. Winlock) postulates two 

64 Ryholt, Situation, 298. 

65 Franke, "Chronologie II", 259. 

66 Ryholt, Situation, 302. 

67 Ryholt, Situation, 193f. For the dates in pBoulaq 18 and pBrooklyn 35.1446 see 
also Franke, "Chronologie II", 254f. and esp. S. Quirke, The Administration of i 
the Late Middle Kingdom. The Hieratic Documents (New Maiden, 1990). 

68 Beckerath, Untersuchungen, 220f. 

69 Franke, "Chronologie II", 265; table, 267ff. 

70 Kitchen (n. 56), 45. 

71 Ryholt, Situation, 195-196. 
' 2 Beckerath, Untersuchungen, 194f.; Ryholt, Situation, 151. 


Theban dynasties, "one prior and one subsequent to the Hyksos con- 
quest of the south". He assignes all the aforementioned 15 places of 
the TC to his first Theban dynasty, which he terms "Dyn. 16", cor- 
recting an error in the terminology of Africanus and traditional research 
(where Dyn. 16 were considered princes and vassals of Dyn. 15), whereas 
the designation "Dyn. 17" applies exclusively to his second Theban 
dynasty.' 3 This "Dyn. 17" would have been listed in a now lost col- 
umn of the TC, being separated from "Dyn. 16" by an alleged local 
dynasty in Abydos (TC XI, 16-3 1). 74 

That allowance must be made for more space after the end of Dyn. 
13 for kings of the Thebaid than is permitted in the traditional 15 
places is an indisputable fact which means that Ryholt's revision is a 
welcome attempt. It is, however, reasonable to suppose that the cor- 
rect reconstruction could have a different appearance. According to 
Beckerath, 7 ' 1 a column might have been lost after TC X (with the first 
entry being TC X, 31), and this column would have included all of 
the Theban kings as a single dynasty; only then would one have to 
add the rulers from Abydos. Beckerath proposes that all of the Theban 
kings between Dyns. 13 and 18 be termed "Dyn. 16", and that the 
local dynasty at Abydos be termed "Dyn. 17". A. Spalinger' 6 is also 
opposed to assuming two Theban dynasties, divided by the postulated 
Hyksos conquest of Thebes. 

Ryholt concludes that the number of Theban rulers was 15 ("Dynasty 
16", listed in the TC) + about 10 ("Dynasty 17", known from the 
monuments and lasting a maximum of ca. 40 years). Taking Beckerath's 
version of Ryholt's reconstruction, the succession of Theban rulers would 
be from TC X, 31 through Beckerath's new column "Xa" to TCXI, 14 
(with the total in TC XI, 15), which would thus include 1 + 31 + 14 = 
46 places, nearly twice as many as in Ryholt's approach. This is not 
conceivable for chronological reasons and with regard to the extant 
epigraphical evidence. 

73 King Sekhemre'-wahkhaw Ra'hotep, traditionally viewed as second king of Dyn. 
17, occupies the first place in Ryholt's Dyn. 17 (Dyn. 17V1 vs. Dyn. 17 T /2) and is 
thus moved 14 places away. Sobekemsaf (traditionally I, Ryholt: II) has been shifted 
18 places. 

74 Ryholt, Situation, 164. 

75 Beckerath (n. 3), 434. 

7b Spalinger (n. 3), 298: "There is little evidence for a Theban dynasty a la Ryholt 
unless we reinterprete the data to suit the hypothesis." 


Regardless of the judgment of the order in the TC, we must con- 
cur with Ryholt that the total number of Theban rulers between Dyns. 
13 and 18 cannot be higher than 25. However, a division of the 
sequence into two dynasties based upon Ryholt's postulated Dyn. 15 
Hyksos conquest of Thebes is just as difficult to accept as Ryholt's pro- 
posed scorched earth policy during the withdrawal of Apopis from the 
south. 77 It is entirely possible that there was originally a division of two 
groups of kings in the TC, but this may have resulted from other con- 
siderations (cf below). These 25 kings at the most seem not to have 
ruled longer than 90—115 years. 

4.2. Dyn. 17 (Ryholt: "Dyn. 16" & "Dyn. 17")— Details The issue of 
ordering these kings is extremely complex due to the state of preser- 
vation of the TC and the epigraphic situation. Even for the more 
important rulers and their families, the data is fragmentary and open 
to diverse interpretations. Three problems of the interconnected fami- 
lies of Inyotef/Sobekemsaf should suffice as exemplary in prohibiting 
a definitive resolution of the issue. 

1) Inyotef Nebukheperre°s consort Sobekemsaf (N.B. "Sobek is his 
(!) protection"), presumably the daughter (or granddaughter?) of a king, 
seems to have been named after a king Sobekemsaf; and the name 
Sobekemsaf was also borne by Ra'hotep's consort (or mother?). 715 Both 
the identity of the Queen(s) Sobekemsaf and the identity of the king 
Sobekemsaf are subject to debate. 

2) Whether this king was the father of Inyotef Nebukheperre c or 
merely an indirect predecessor depends upon the interpretation of a 
newly found fragmentary inscription on the Luxor-Farshut road. 79 

3) The precise placement of the Inyotef kings, the Sobekemsaf kings 
and the king RaTiotep is subject to controversy. That a prince with 
the basilophorous names Inyotefmose was praised by a king Sobekemsaf 
for his actions during a Sokar festival could demand at the most the 
placing of a king Sobekemsaf after the Inyotef kings. 80 

The following brief presentation attempts to provide a solution along 
this line of arguments: A detailed discussion must take account of the 

77 Ryholt, Situation, 143-148. 

78 Ryholt, Situation, 265-266, 268ff. 

79 J. C. Darnell & D. Darnell, "The Luxor-Farshut Desert Road Survey", Oriental 
Institute Annual Report (1992-93), 50, fig. 4; Ryholt, Situation, 270. 

80 Ryholt, Situation, 170. 


aforementioned differences between Ryholt's interpretation of the TC 
and Beckerath's response to it (cf. above, with Figure). One possible 
solution is to reject the proposed additional column TC Xa (with Ryholt, 
against Beckerath), while simultaneously assuming (with J. P. Allen, 
against Ryholt) 81 that all of the Theban rulers between Dyns. 13 and 
18 were originally listed between TC X, 31 and TC XI, 25. A first 
group (TC X, 31 —XI, 14) would close with the total at TC XI, 15; and 
then 9 additional rulers would follow in TC XI, 1 6-XI, 24 whose total 
would once have been noted in TC XI, 25 (before the inception of fr. 
163). 82 The division into two groups could easily have been motivated 
by their lineage or residence (e.g., Thebes vs. Dendera). 83 

The five kings mentioned on fr. 163 would only be inserted then, 
and possibly others on a possible column "XII" . That these will have 
been Abydenian rulers — Ryholt suggests that the remainder of his 
Abydos dynasty was listed here — is not demonstrable since the traces 
of the names do not match any epigraphically attested names. 

Despite diverging on the reconstruction of TC XI, I agree with Ryholt 
in the number of rulers who can be viewed as Theban kings between 
Dyns. 13 and 18. In addition to the 15 places given in the TC (i.e., 
the traditional Dyn. 17), Ryholt names 9 kings of a "second Theban 
dynasty" (i.e. his "Dyn. 1 7"). According to the placement of these rulers 
in TC XI such as proposed above, there would in fact be exactly that 
many places. 

The first three places of TC XI are damaged and begin with Shm- 
/-, Shm-f- and Shm-f-s-. The traditional restoration (since H. Stock) to 
the throne names Shm-r c -w>h-h c w (— Ra'hotep), Shm-r c -w>d-h c w 
(= Sobekemsaf I) and Shm-r' '-smn-tiwj (= Djehuti) has recently been 
thrown into doubt by Beckerath himself, who considers Ra'hotep as 
doubtful, the third place as completely uncertain, and views only 
Sobekemsaf I as certain. 84 On account of their building activity in 
Abydos and their restorations in Medamud and Coptos, Ryholt places 
Ra'hotep and Sobekemsaf Sekhemre'-wadjkhaw in his chronologically 

81 Allen in: Ben-Tor et al. (n. 3), 49. 

82 An alternative placing of the fragments is not possible; cf. W. Helck, "Anmerkungen 
zum Turiner Konigspapyrus", SAK 19 (1992), 151-216. 

83 Cf. below, text referring to n. 111. 

84 J. v. Beckerath, "Theban Seventeenth Dynasty", in: E. Teeter & J. A. Larson, 
eds., Gold of Praise. Studies on Ancient Egypt in Honor of Edward F. Wente (Chicago: SAOC 
58, 1999), 23-24. The earlier sequence can still be found in Beckerath, Chronologie, 139. 



Franke, "Chronologic" II (1988) 

Ryholt, Situation (1997) 

DYN. 17 

I) Ncbukhcpcrrc / Inyotcf V 

2} Sekhcmrewahkhau / Raholep 

3) Sekhemrewadjkbau / 
Sobekemsaf I 

4) Sekhemre / Sementawy / Djehuti 

5) Seanklienre / Mentuhotep VI 

6) Sewadjenrc / Nebiriau (I) 

7) *Ncbiriau (II) (mistake?) 

8) Semenre 

9) Seuserenre / Bebiankh 

10) Sekhemreshedtawy/waset/ 
Sobekemsaf II 

I I) Sekhemre up imaat / Inyotef VI 

12) Sckhcmrchcrhcrmaat / Inyotcf VII 

13) Senakhtenrc 

14) Seqenenre / Tao 

15) Wadjkheperre / Kamose 

Dyn. 1 6 = Dyn. 18,1, Nebpehtire / Ahmose 

Unplaced kings, Thebes: 
Senwosret IV Seneferibre 
Mentuhotep V Merankhre 

Neferhotep III Sekhemreseankhtawy / 
Sobekhotep VIII Sekhemreseusertawy 

Dyn. 16- r r tiiobanDyn. 

2) Djehuti Sekhemre sementawy 

3) Sobekhotep VIII 

4) Neferhotep III 

5) Mentuhetepi Seankhenre 

6) Nebiriau I Sewadjenre 

7) Nebiriau II- 

8) - Semenre 

9) Bebiankh Seuserenre 

10) - Sekhemreshedwaset 

1 1-15) lost (TC) 

Dyn. 17 - 2 nd TIIEBAN DYN. 

1) Rahotep Sekhcmrewahkhau 

2) Sobekemsaf" I 

3) Inyotef VI Sekhemreup imaat 

4) Inyotef VII Nebukheperre 

5) Inyotef VIII Sekhemreherhermaal 

6) Sobekemsaf II 
Sekhemrewadj khau 

7) Siamun (?) Senakhtenre 

8) Tao Seqenenre 

9) Kamose Wadjkhcpcrrc 

(Dyn. 18/ 1) Nebpehtire/Ahmose 

Unplaced kings, Thebes (La..): 
Dedumose I Djedhetepre 
Dedumose II Djedneferre 
Mentuemsaf Djedankhre 
Mentuhotep VI. Merankhre 
Scsostris IV Seneferibre 

Fig. II. 7.2 


later "Dyn. 17" 8 '' and restores TC XI, 1—3 with the throne names Shm- 
r c -smn-t>wj (= Djehuti), Shm-r c -swsr-t>wj {— Sobekhotep VIII) und Shm-r c - 
s'nh-tiwj (= Neferhotep III), of whom the last two were traditionally 
acknowledged as contemporary but had not been placed. 86 Djehuti is 
thus shifted only two places in contrast to the traditional sequence (Dyn. 
17 T /4 and "Dyn. 16" R /2). 87 Placing Neferhotep III here is favoured by 
the position of Se c ankhenre c Mentuhotpi (Mentuhotep VI) in TC XI, 
4, as two similarly executed stelae of the two kings could possibly have 
come from the same workshop. 88 The next five places TC XI, 5—9 are 
preserved whereby TC XI, 6 (Nebiriaw II) could be a copying error 89 
and in TC XI, 9 Sekhemre'-shedwaset (as traditionally maintained) 
might be a deliberate change for the correct Sekhemre'-shedtawy = 
Sobekemsaf [I]. 90 The proper name of TC XI, 7, Semenenre c , is un- 
known; as the successor is named "Bebiankh" and "Bebi" is a possi- 
ble abbreviation of a name containing the element "Sobek", 91 one could 
speculate on a Sobek-name. Of the following five places (TCXI, 10—14), 
only the insignificant beginning is preserved, before the total (TC XI, 
15). TCXI, 15-17 apparently introduces a new group of kings. 

The main difficulty here is the correct restoration of the five places 
before the sum. If one places the traditional Dyn. 17 in TC X, 31— XI, 
14, the last three places are occupied by Senakhtenre c , Seqenenre c Tao 
and Wadjkheperre c Kamose. 92 Ryholt assigns to these positions the 

85 Ryholt, Situation, 152, 170. Caution is, however, advised in basing the restoration 
of Sobekemsaf Sekhemre'-wadjkhaw on the alleged destruction of the Thebaid by the 
Hyksos (145-146; 170). 

86 N. Dautzenberg, "Neferhotep III. und Sebekhotep VIIL", CM 140 (1994), 19-25, 
also views the second as being the immediate successor, but places them in Dyn. 13. 
The same restoration for TC XI, 2 was accomplished by C. Bennett, "The First Three 
Sekhemre Kings of the Seventeenth Dynasty", GM 143 (1994), 21-28. 

87 A connection to Dyn. 1 3 follows via the vizier Ibiaw, for which link one should note 
the critical remarks by Grajetzki (n. 3), 151-152. Franke, Heqaib, 79 is more positive. 

88 P. Vernus, "La stele du roi Sekhemsankhtaouyre Neferhotep Iykhernofret (Stele 
Caire JE 59635) et la domination Hyksos", ASAE 68 (1982), 129-135 with pi. I; idem, 
"La stele du pharaon Mntw-htpi a Karnak: Un nouveau temoignage sur la situation 
politique et militaire au debut de la D.P.I.", RdE 40 (1989), 145-161, pls.6-7; Ryholt, 
Situation, 154. Beckerath, Chronologie, 139, places Senwosret IV Seneferibre' at TCXI, 4. 

89 Franke, "Chronologie II", 263. 

90 Following Beckerath, Untersuchungen, 168, 290, contra Ryholt, Situation, 156. 

91 P. Vernus, Le surnom au Moyen Empire (Rome: Studia Pohl 13, 1986), 111 n. 122; 
cf the two sons of Sobekhotep VII Merkawre' with the names Bebi and Sobekhotep 
(Ryholt, Situation, 235-236). 

92 Franke, "Chronologie II", 271. 


rulers Dedumose I, Dedumose II, Mentuemsaf, Mentuhotep (VI R ) 
Mer c ankhre c and Senwosret IV Seneferibre c , who are documented epi- 
graphically at Thebes but had not been positioned chronologically. 
However, the Dedumose kings certainly belong in Dyn. 13. 93 If one 
follows the scheme outlined above, with the assumption that further 
Theban kings down to Kamose were named in TC XI, 16— XI, 25, we 
face the question of the placement of the kings Inyotef/Sobekemsaf/ 
Ra'hotep and thus of the exact relationship between the late Dyn. 13 
and the kings named between TC X, 31 and TC XI, 25. 

As Ryholt correctly notes, the traditional placing of Nebukheperre c 
Inyotef (V) at the beginning of Dyn. 17 (his "Dyn. 16"), in the lost 
entry at TC X, 31, cannot be justified. D. Polz, who succeeded in 
finding the tomb of the king in Dra Abu'l Naga in 2001, argues like- 
wise for situating him before the final kings of Dyn. 17. 94 Such a place- 
ment is likewise favoured by the evidence of the box of Minemhat, 
who was mayor of Coptos under Inyotef V, which was part of the 
funerary equipment of an Aqher who lived under Seqenenre'. 93 

From the legend on the coffin Louvre E 3019, it follows that Inyotef 
Nebukheperre c (Dyn. 1 7 T / 1 : Inyotef V; Dyn. 1 7 R /4: Inyotef VII) arranged 
the burial of his brother Inyotef Sekhemre'-upimaat (Dyn. 17 T /11: 
Inyotef VI; Dyn. 1 7 R /3) and must therefore have followed him on the 
throne. 96 In his Untersuchungen, Beckerath had viewed Inyotef Sekhemre c - 
upimaat (VI) and Inyotef Sekhemre'-herhermaat (VII) as brothers, 
whereas he had separated Inyotef Nebukheperre c (VI; coffin BM 6652) 
from them as a king he considered not necessarily related to them, 
placing him at the beginning of the dynasty. Ryholt equally bases his 
arguments upon a consistent palaeographic peculiarity (the Pleneschrei- 
bung of "j") in the case of the coffin of Inyotef Sekhemre'-herhermaat 

93 Franke, Heqaib, I'i '-78. 

94 D. Polz & A. Seiler, Die Pyramidenanlage des Konigs Nub-Cheper-Re Intef in Dra' Abu 
el-Naga. Ein Vorbericht (Maniz: DAIKS 24, 2003). J. C. Darnell views Inyotef Nebukheperre' 
as the addressee of a newly discovered royal hymn at Wadi el-H61; however, the king 
is not named ("A New Middle Egyptian Literary Text from the Wadi el-H61", JARCE 
34 [1997], 85-100). 

95 H. Winlock, JEA 10 (1924), 258 with n. 1 (taking up an idea of P. Newberry); 
W. Helck, "Der Aufstand des Tetian", SAK 13 (1986), 126 (who erroneously refers to 
a spoon; the spoon from the burial of Aqhor came from a mayor Sobekwer). In favour 
of a placement in late Dyn. 1 7 (as traditionnally conceived) is N. Dautzenberg, "Die 
Wahl des Konigsnamens in der Hyksoszeit", GM 159 (1997), 43—52, based on the 

96 Ryholt, Situation, 270. 


(Dyn. 17V 12: Inyotef VII; Dyn. 17 R /5: Inyotef VIII; Louvre E 3020) 
so that he becomes an ephemeral coregent of Inyotef Nebukheperre 1 . 97 
However, this "stock coffin" was certainly not originally intended for 
Inyotef Nebukheperre c , and to infer from it any hypothetical coregency 
is doubtful in the extreme. 9 " 

Placing the three Inyotef kings together appears to be plausible, not 
the least on account of their canopic equipment. 99 It seems possible 
that the legend on Louvre E 3020 can be used to argue that Inyotef 
Sekhemre'-herhermaat (VIII) was the son of Inyotef (VI) Sekhemre c - 
upimaat, 100 which would produce the sequence: Inyotef (VI) Sekhemre c - 
upimaat ("the elder") — Inyotef (VII) Nebukheperre c — Inyotef (VIII) 

If we can identify Sobekemsaf, the consort of Nebukheperre c with 
the mother of Ra'hotep (presumably) of the same name, 101 then Ra'hotep 
would have to be placed chronologically after Nebukheperre c . 102 The 
evidence is of similar ambiguity with regard to the mutual position of 
Ra'hotep and Sobekemsaf Sekhemre'-wadjkhaw. If a prince Ameni 
whom Ra'hotep grants a bow, arrows and the right to participate in 
the ceremonies for Min of Coptos, is identical to the son-in-law of 
Sobekemsaf Sekhemre'-wadjkhaw, then Ra'hotep would plausibly be "a 
close predecessor of Sechemre'-wadjkhau Sobekemsaf". 103 If the two are 

Cf. Ryholt, Situation, 267-268.; the earlier discussions will be found in Beckerath, 

98 Dodson (n. 3), 50-51; Spalinger (n. 3), 300. 

99 A. Dodson, The Canopic Equipment of the Kings of Egypt (London, 1994), 42; C. Bennett, 
"The Date of Nubkheperre Inyotef", GM 147 (1995), 19-27 (22f: all three Inyotefs 
closely associated). 

100 The coffin was not originally manufactured for Inyotef Nebukheperre' (A. Dodson). 
When the coffin was reworked, the <y> of the name of Inyotef was replaced with 
<','>. An explanation can be found in assuming that the correction was an attempt at 
specification: the proper name Inyotef, "The one who brings back his father" is an 
Ersatzname (a newborn child is believed to replace a recently deceased relative, in this 
case the father). The insertion of <''> and the addition of the throne name would 
have adapted the meaning of the name to the contemporary situation: "he who brings 
back the elder, his father, Sekhemre'-herhermaat". As the epithet <'>> "the elder" is 
only attested for Sekhemre'-upimaat, it would follow that Inyotef Sekhemre-herher- 
maat was the son of Inyotef Sekhemre'-upimaat. 

101 For the discussion, cf. Ryholt, Situation, 265-268 (who rejects the identification). 

102 Ryholt, Situation, 170, opposes this order: "Likewise, it may be argued that 
Ra'hotep, whose restoration of the temple of Min at Coptos is described on a stela, 
was a predecessor of the Inyotef kings, since such a claim could hardly have been 
made in the years subsequent to the reign of Inyotef N[ebucheperre'] who built exten- 
sively at this temple." 

103 C. Bennett, "The First Three Sekhemre Kings of the Seventeenth Dynasty", GM 
143 (1994), 21-28. 


not the same, 104 then the chronological place of Sobekemsaf Sekhemre c - 
wadjkhaw is completely uncertain; rather than placing him, with Ryholt, 
after the Inyotef kings, 1(to he could also precede them, perhaps after 
Sekhemre'-shedtawy Sobekemsaf (I) or even in the lost entry at TC X, 
31. 106 A possible candidate for TC X, 31 or one of the other open 
places could, however, also be seen in Sekhemre'-neferkhaw Wepwaw- 
temsaf (previously unplaced, assigned by Ryholt to the Abydos dynasty). 
In principle, the block of the Inyotef kings and Ra'hotep could have 
stood in TC XI, 10-14 or else in TC XI, 18-21 (with the four fol- 
lowing entries to be restored as Senakhtenre/SeqenenreVKamose/total). 
In view of these possibilities, of particular interest are the two entries 
where the beginning is still preserved in TC XI, 16 and TC XI, 17, 
which Ryholt views as undocumented throne names of kings of his pos- 
tulated Abydos dynasty. In TC XI, 16 Weser-. . . / is preserved, and in 
TC XI, 17 probably only Wsr without the solar disk, 10 ' i.e., there is a 
proper name as in TC VIII (Ryholt: IX), 1 (Msj), TC XI, 5/6 Nb-jrj- 
>w, and elsewhere. 108 For both places, which have not hitherto been 
explained, a restoration can be suggested: 

• In the tablet of kings from Karnak, the fourth row (no. 27) has a 
royal name Wsr-n-r c along with Seqenenre c , Senakhtenre c and Inyotef 
Nebukheperre c , who has usually been emended to the Swsr-n-f named 
in TC XI, 8. 109 However, the kings to be placed in TC X, 31— XI, 
14 (including Swsr-n-f) are certainly listed in other rows of the tablet 
of kings if the assignment of kings to the TC passage as proposed 
above can be accepted. 

• A king who fits well with the notation beginning Wsr- at TC XI, 1 7 
is the hitherto unsituated Senwosret (IV) Seneferibre c , known from 
a colossal statue and a stela in Karnak as well as blocks from Tod 
and Edfu. 110 

104 Ryholt, Situation, 266, 272 (the names display different endings: Amanja, Amana). 

105 Decisive for Ryholt is the praise of a king Sobekemsaf for a Inyotefmose, whom 
Ryholt places after the Inyotef kings due to his basilophorous name. 

106 Franke, Heqaib, 84: the dyad of king Sobekemsaf Sekhemre'-wadjkhaw with Satet 
is characteristic for the beginning of Dyn. 1 7 . 

107 The weser-sign follows on the cartouche more closely than in TC XI, 16. 

108 In detail: Ryholt, Situation, 27-28. 

109 Beckerath, Untersuchungen, 27; idem, Handbuch, 126. It is less probable that the Beni 
Hasan graffito (cf. Beckerath, Untersuchungen, 69) should be read as Weserre'f-nefer- 
khaw] . 

110 Ryholt, Situation, 391 (16/e), 157, 306; Beckerath, Untersuchungen, 255 (13 F), 62. 


In the lacuna between TC XI, 17 and XI, 26, I would propose plac- 
ing the kings Senakhtenre c , Seqenenre c Tao, Wadjkheperre c Kamose 
(and possibly others) as well as the total. The rulers named in the lower 
part of TC XI would thus be the kings of Ryholt's "Dyn. 17" (whom 
he believes to have been listed on a lost * twelfth column). Their sep- 
aration from those named in the upper part of the column could have 
been motivated by this family's suggested provenance from Dendara 
and its struggle with the supplanted Inyotef family that continued until 
early Dyn. 18. 111 

Frag. 163 placed at the end of TC XI has a throne name ending 
in — h'b in line 2 6, and another in line 31 might be constructed with 
wbn. A parallel can be found in TC VIII, 4 (Shib-/) and TC VIII, 1 1 
(Wbn-f), whereas there are absolutely no corresponding throne names 
in Dyns. 17-19. 112 

What follows is a listing of TC X, 31— XI, 26 with the hypothetical 
restorations proposed above in the right column: 

TC X, 



Upuautemsaf ?? 




Sementawy Djehuti 




Seusertawy Sobekhotep VIII 




-'anchtawy Neferhotep III 







Nebiriaw (I.) Sewadjenre' 



Nebiriaw (II.) 




(proper name unknown; 
Sobek- ?) 



Seweserenre' Bebiankh 



Sekhemre'-shedwaset (< tawy ?) 

Sobekemsaf (I) 



Sobekemsaf II 



Sekhemre'-upimaat ("the 
elder") Inyotef (VI) 

[continued on next page) 

Cf. also Beckerath's "Sesostris V". Improbable seems Weser[monthu] (?) (Beckerath, 
Untersuchungen, XIII, L). 

111 Helck (n. 95), 125-133; cf. the Coptos decree of Inyotef (VIII, according to the 
sequence presented here): E. Martin-Pardey, "Zum Koptosdekret Antefs V.", in: FS 
Jiirgen von Beckerath zum 70. Geburtstag am 19. Februar 1990 (Hildesheim: HAB 30, 1990), 


112 From late Dyn. 17 there is only a titulary prince Sbk-m-hb attested in Esna; and 
the consort Sbk-m-hb of a prince Ameni. The next ruler with a name (but not the 
throne name) ending with -m-hb is Horemhab. 


table {cont) 

TC XI, 1 1 Sekhemre'-Herhermaat 

Inyotef (VII) 

TC XI, 12 Nebukheperre 1 Inyotef (VIII) 

TC XI, 14 Sekhemre'-wahkhaw 


TC XI, 15 [1]5 kings 

TC XI, 16 Weser ... re' Weserenre' 

TC XI, 17 Weser- Senwosret (IV) Seneferibre' 

TC XI, 18 

TC XL, 19 [placing of the following 

kings uncertain] 

TC XL, 20 Senakhtenre' 

TC XL, 21 Seqenenre' Tao 

TC XI, 22 Wadjkheperre' Kamose 

TC XI, 23 (total) 

TC XI, 24 

TC XI, 25 

TC XI, 26 ... hab 

The lengths of the reigns are listed in the TC as follows: TC XI, 1: 3 
years (rest lost); TC XI, 2: 16 years (rest lost); TC XI, 3: 1 year (rest 
lost); TC XI, 4: 1 year (rest lost); TC XI, 5: 26 years (rest lost); TC 
XI, 8: 12 years, x months, 12 days. 113 Of those rulers who are hypo- 
thetically arranged here among their successors, a regnal year 7 is 
attested for Sobekemsaf II, and for Inyotef VIII (Ryholt: Inyotef VII) 
Nebukheperre' a year 3. These eight kings produce a total of 69 years; 
for all 15 entries TC X, 31— XI, 14 we might estimate 75—100 years. 114 
For the kings placed from TC XI, 16 onward, there are good reasons 
for assigning them a significantly shorter length of rule. Kamose will 
have died not long after his regnal year 3 (attested on the Kamose 
Stela), at a time when Ahmose was still quite young. 115 As his mummy 
testifies, Seqenenre' suffered an early death on the battlefield. Ryholt 
postulates 4 regnal years for Seqenenre', and one year for Senakhtenre 1 , 116 
who is not recorded in contemporary documents. If Senwosret IV is 

Cf. Ryholt, Situation, 202. 

Cf. — with a different succession — the estimates by Ryholt, Situation, 204. 
For the coregency, cf. Ryholt, Situation, 273; otherwise, 172ft, 309. 
Ryholt, Situation, 206. 


to be placed here, a first year is attested in a biographical inscription. 117 
It follows that the rulers listed here were in power only very briefly; 
with a total of perhaps 15 years. 

Despite various differences in the internal sequence of kings between 
Ryholt's version and that presented here, both reach a relatively reli- 
able estimate of the length of time to be assigned to the rulers between 
Dyn. 13 and 18: roughly a century. Ryholt has 67 years ("Dyn. 16") 
+ 31 years ("Dyn. 17"). According to our reading TC X, 31— XI, 14 
covers 75—100 years, and TC XI, 16 to Kamose amounts to 15 years 
(cf. below, final remarks, section 6). 

5. Dynastie 15 

The kings of Dyn. 15, or the "Hyksos" (Hkl.w-hls.wt, "Rulers of Foreign 
Countries") were rulers of (at least remote) Syro-Palestinian descent 
whose power was based in the eastern Nile Delta where a high pro- 
portion of the population were of Palestinian origin. Establishing the 
relative chronology of this dynasty suffered from the inadequate docu- 
mentation and the widely assumed discrepancy between the royal names 
in the epigraphic sources and the Manethonian tradition. 11 " Previously, 
moving beyond the Khamudi attested in the TC and the epigraphi- 
cally recorded kings involved deciding which of the many scarabs named 
kings of Dyn. 15 and which did not. Hitherto, there were no gener- 
ally recognized criteria for the attributions. The distribution and fre- 
quency of their scarabs often seemed to favour including Sheshi and 
Ja'qubhaddu ("Jaqobher") whereas Helck believed that "Semqen" and 
'Anathaddu ('"Anather") could be assigned a place among the "great Hyksos" 
(i.e. Dyn. 15 of the traditional approach). Ryholt proposed to include 
the Skr-hr now attested in Tell el-Daba and to consider, from the three 
rulers whose scarabs include the title hql-hlswt, Smkn and c pr- c nt as the 
first two kings of Dyn. 15. 119 None of these names can be equated with 
the Hyksos names such as preserved by Manetho, and this alleged dis- 

"' W. Helck, Historisch-biographische Texte der 2. Zwischenztit und neue Texte der 18. Dynastie 
(Wiesbaden: KAT, 1983 2 ), 41 [no. 56]. 

us p or th e history of the debate from 1936—1997, see in detail, Schneider, Auslander, 

119 Ryholt, Situation, 118—125. Moving the third — 'nt-hr — to Dyn. 12 is apparendy 
not correct, cf. Ben-Tor et al. (n. 3), 63. 


crepancy seemed to indicate to most scholars that the Manethonian 
evidence has to be discarded despite of the close to complete lack of 
contemporary sources that might fill the gap. The situation was that 
summarized by W. A. Ward: "It is impossible to equate the names preserved 
in the various recensions of Manetho with these actually known from the monu- 
ments (. . .) It is clear that most of the names preserved in this tradition are too 
corrupted to have any value (. . .) The present discussion will therefore ignore 
Manetho as being unreliable." 1 ' 20 

This traditional standpoint relies on the improper assumption that 
even if most of the documentary evidence on the Hyksos is lost, we 
nevertheless possess at least all their names, and does not accurately 
reflect on how their names were handed down and copied in the later 
chronographical tradition. Instead of rejecting Manetho, the author has 
tried 121 to solve the issue by reckoning with kings absent from our con- 
temporary documentation 122 and by accounting for textual mistakes in 
the process of the copying of the king lists by late scribes who could 
solely rely on the written form of the names but were not aware of 
their original sounding. Our correlation of the traditions does not leave 
any lacunae which must be filled having recourse to scarabs. The nearly 
complete loss of the sequence of Dy. 15 kings in TC (where only the 
last one, Khamudi is preserved) gives priority to Manetho of whose 
epitomists those can be shown to be correct that place Apophis at the 
end of the dynasty (Africanus, Eusebius's Armenian version, the scho- 
lion to Plato's Timaios). The names of this sequence — 1. Salitis, 2. 
Bnon, 3. Apachnan, 4. Iannas, 5. Archles/Assis, 6. Apophis — can all 
be equated with names attested epigraphically with one exception for 
which a postulated original name can be supplied. 123 It has to be noted, 

120 W. A. Ward, "Royal-Name Scarabs", in: Studies on Scarab Seals. Vol. II. Scarab Seals 
and their Contribution to History in the Early Second Millennium BC (Warminster, 1984), 162. 
Similarly, M. Bietak, Historische und archaologische Einfuhrung, in: Pharaonen und 
Fremde. Dynastien im Dunkel. Sonderausstellung des Historischen Museums der Stadt Wien in 
Zusammenarbeit mil dem Agyptologischen Institut der Universitdt Wien und dem Osterreichischen 
Archaologischen Institut Kairo, Rathaus Wien, Volkshalle, 8. Sept.-23. Okt. 1994, 17-57: 24. 

121 Schneider, Auslander, 33-56.70-75. 

122 Striking examples are the two Hyksos Skr-Hr (architrave found in Tell el-Daba 
and first published in 1994) and Khamudi (attested only in TC) who could not possi- 
bly have been postulated by modern research. A late tradition of Skr-Hr seems to be 
preserved in pCarlsberg 642 where an impious ruler Saker is mentioned (cf J. F. Quack, 
"Zwischen Sonne und Mond — Zeitrechnung im Alien Agypten", in: H. Falk, ed., Vom 
Herrscher zur Dynastie. Z um Wesen kontinuierlicher J^eitrechnung in Antike und Gegenwart, Bremen 
2002, 27-67: 47f), equated by Quack with Salitis, but Skr-Hr is certainly preferable. 

123 For what follows see in detail Schneider, Auslander, 50-56. 



that the TC apparently once listed a total of 6 rulers of this dynasty, 
including one named Khamudi, the copyists of Manetho also name 6 
rulers, but without Khamudi. 124 


Copyists of Manetho: 

a. Sara-Dagan (Shk[nJ) 

1 . Salitis 

> 36 years, 7 months 

b. *Bin-'Anu 

2. Bnon 

> 44 years 

c. ( 3 Apaq-) Hajran 

3. Apakhnan 

> 19 years 

d. Jinassi'-Ad 

4. Iannas 

> 50 years, 1 month 

e. Sikru-Haddu 12 ' 

5. Archies /Assis 

> 49 years 2 months 

f. Apapi 

6. Apophis 

> 61 years 

6. Halmu'di 

— (not in Manetho) 

Total TC: 108 years 




The total given at TC X, 21 for the entire Dyn. 15 is 108 years. 126 By 
contrast, according to Africanus's version of Manetho, the dynasty would 
have lasted 284 years. Assigning reign lengths to individual rulers is at 
the present possible only in some cases. Beckerath proposed a hypo- 
thetical reign for all individual rulers, 127 but this remains speculative 
since it is not clear which of the reigns have been lengthened by decades 
by Manetho and which not. On an unnumbered fragment of the TC 
(Ryholt's "b") 128 are the remains of the totals of two reigns: "10 (or 
20, or 30)" + x years and "40" + x years. The latter could hypo- 
thetically be assigned to Apapi, who according to the colophon on the 
Rhind Mathematical Papyrus must have reigned at least 33 years. The 
first reign might then be assigned to Skr-Hr. m Another reign length 
can be inferred from the note on the verso of the Rhind Mathematical 

124 Cf. Schneider, Auslander, 56 (if not in the name Chamois of the book of Sothis). 

125 The interpretation of major elements of these names such as offered by Ryholt, 
Situation, 126ff. (particularly 'pr and hr) reflects older positions (Albright) which are no 
longer supported. 

126 See Ryholt, Situation, 118. — It must be noted at the outset, however, that in an 
unpublished communication made while this book was in its final stages of prepara- 
tion, Ryholt announced that the traces on the actual papyrus did not confirm Gardiner's 
reading and that another date, possibly of more than 140 years should instead be con- 
sidered possible (editors' note). 

127 Beckerath, Chronologie, 137. 

128 Ryholt, Situation, 118-119. 

129 Contra Ryholt, Situation, 119. 


Papyrus whereby in the 11th regnal year of the ruling king, Heliopolis 
has been conquered, and "he of the South" has attacked and taken 
Sile. Since "he of the South" must denote the Theban ruler Ahmose, 
the regnal year 1 1 can only be assigned to the successor of the Hyksos 
king Apapi: Khamudi. 130 The Hyksos capital Avaris will have fallen to 
Ahmose not much later. An inscribed spear point from the booty of 
Avaris which includes a dating criterion (orientation of the lunar hiero- 
glyph) 131 enables us to place the conquest of Avaris in the regnal year 
18/19 of Ahmose. A possible check is provided by a graffito in the 
quarry at Tura whereby "oxen from Palestine" were used at the open- 
ing of the quarry in Ahmose's regnal year 22. 132 The cattle could have 
been brought to Egypt after the three-year siege of south Palestinian 
Sharuhen, which followed the conquest of Avaris. This link between 
the reigns of Khamudi and Ahmose means that the beginning of Dyn. 
15 should be 90 years before the death of Kamose (108 years — pro- 
vided this number from the TC is correct — minus the 1 8 years of the 
reign of Ahmose parallel with the last two Hyksos). 

6. Total Length of the Period of Dyns. 13—17 

Providing any exact figure for the total duration of the period of Dyn. 
13-17 is impossible. The difficulties are not merely due to the inade- 
quacy of the documentation of the various dynasties, but rather that 
these are exacerbated by our inability to establish their mutual rela- 
tionships. In particular, it is far from certain that Dyn. 15 and the 
Theban dynasty (Ryholt's "Dyns. 16/17") followed immediately on Dyn. 
13 as Ryholt proposes, with both starting in 1649 BC. It seems plau- 
sible that the transition was both spatially and temporally more com- 
plex in the final decades of the 1 7th century. The fact that the numerous 
ephemeral rulers of the last 25 years or so of Dyn. 13 were probably 
incapable of maintaining administrative control of the entire country 

130 D. Franke, "Chronologie II", 263; M. Bietak, "Historische und archaologische 
Einfuhrung", in: Bietak, SCIEM 1996/ 1998, 29; Beckerath, Chronologie JVR, 115; Spalinger 
(n. 3), 299. Ryholt is opposed, but procedes a priori from a short reign for the last 
Hyksos king; likewise, Kitchen, "Regnal and Genealogical Data of Ancient Egypt," in: 
Bietak, SCIEM 1996/1998, 39-52, esp. 45-46 (who also considers assigning the date 
to the last Hyksos). 

131 Kitchen (n. 130), 46 correctly stresses that this criterion is not well founded. 

132 Franke, "Chronologie II", 264. 


suggests that there may have been a number of competing local dynas- 
ties at the time. 133 It would be plausible to suppose that the origins of 
the power of these local dynasts lie in the period before the ephemeral 
end of Dyn. 13, and therefore that an approximate chronological fixpoint 
might be established at the end of the reign of Aya, at ca. 1650 BC. 134 
This linkage between Dyns. 13 and 17 finds some support in bio- 
graphical information of the time. 135 

Calculating back from the start of the reign of Ahmose (around 1539 
BC) and suggesting that the Theban kings reigned for 90—115 years 
before Ahmose leads to a beginning of the Theban dynasty at ca. 
1654/1629; based on the TC, the dates for Dyn. 15 would be 1639-1521 
BC (cf. above). Given the uncertainties in the length of individual reigns 
here and in Dyn. 13, a conclusive judgment about the length of the 
SIP 136 still needs further evidence. 

133 Spalinger (n. 3), 297-298 assumes that an independent kingdom (= Dyn. 17) 
already appeared a few decades before the end of Dyn. 13, and poses the question of 
the possible coexistence of a truncated state in the North with the Hyksos. 

134 Cf. still the remark of the Greek author Artapanos (2nd century BC), that under 
a king named "Chanephres" (Khaneferre' Sobekhotep IV), Egypt was divided into var- 
ious kingdoms (Schneider, Ausldnder, 158-159). Whether the pyramidion of Aya found 
at Faqus was removed there from Saqqara by the Hyksos, as Ryholt (Situation, 147) 
and Dodson (n. 3; 50) assume is not certain; Beckerath (Untersuchungen, 73) assumed 
the Ramesside era. Of the various possible synchronisations between the dynasties, 
Manetho's remark about a supposed conquest of Egypt by the Hyksos under a 
"Toutimaios", in whom one of the Dedumose kings of Dyn. 13 had earlier been sus- 
pected, is no longer tenable as the reference is clearly due to a misunderstanding of 
the Vorlage, or a textual error: cf. Schneider, Ausldnder, 159; for a different interpreta- 
tion, see A. Biilow-Jacobsen in: Ryholt, Situation, 327ff. 

130 The stele juridique records that an Aya received the governorship of el-Kab in reg- 
nal year 1 of Merhotepre' Ini (successor of the king Aya) and that his grandson Kebsi 
sold it in regnal year 1 of Nebiriaw. In TC XI, 1-5, the lengths of the reigns up to 
Nebiriaw I are preserved (21 years, with a few months missing), so that including the 
reign of the founder of the dynasty in TC X, 31, some 25 years will have passed 
before regnal year 1 of Nebiriaw (cf. above, section 4). If the dynasty began between 
1654 and 1629 BC, Merhotepre' Ini (cf. above, section 3) should be placed around a 
century after the start of the dynasty, i.e., ca. 1656 BC. This would result in 27-52 
years for the period during which the father and grandfather of Kebsi exercised the 
office, which appears plausible. Ryholt (Situation, 197, 202) has a far larger margin of 
more than 70 years (1 Merhotepre' Ini = 1677 BC; 1 Nebiriaw I = 1627 BC), assum- 
ing that the father and grandfather of Kebsi reached a very ripe old age of more than 
70 years each. 

136 Ryholt, Situation, 191 opts for a long, of 254 years (as R. Parker proposed based 
on his absolute dates for the MK), and thus proposes the dates of 1803-1549 (Dyn. 
14 from 1805; Dyn. 15 to 1540; similarly Kitchen (n. 130), 46. For a short chronol- 
ogy cf. E. Hornung ("Lang oder Kurz?," in: High, Middle, Low 1, 36; 1756-1539 BC, 
with Dyn. 15 down to 1521. Dodson (n. 3, 50-51) also concludes that a maximum 
chronology is not necessary, but is willing lower the beginning of the NK to 1500 BC. 


Erik Hornung 

For some time there has been a consensus about the relative chronol- 
ogy of the New Kingdom. Since most reign lengths are well docu- 
mented, they are not problematic. Nor does the sequence of kings pose 
difficulties — except for Akhenaten's immediate successors (who are not 
mentioned in the cultic lists). However, the issue of coregencies for 
Thutmose III/Amenhotep II and Amenhotep III/Akhenaten continues 
to stimulate debate, as does the exact relationship between Amenmesses 
and Sety II, although otherwise dates important for relative chronol- 
ogy are particularly abundant for the Ramesside Period, leaving very 
little "empty space". 

This positive situation is enhanced by the fact that regnal years were 
counted from a king's accession and appear in the dates of documents 
which allows precise calendar dates to be suggested for some reigns, 
the necessary first step leading to a "day-exact chronology" (Depuydt), 
as is in fact possible in the Late Period. The identification of the exact 
day of accession or the establishment of very precise possible limits 
allows additional checks through the months offered by Manetho via 
Josephus. A few problematic issues remain, especially the length of 
Haremhab's reign which has been estimated at between 13 and 27 
years. In this particular case, additional criteria, such as the number 
of monuments or the sequence of officials cannot solve the problem. 

Nor do we have Haremhab's mummy which could have provided a 
potential check based on his estimated age at death. But in fact, age 
estimates for royal mummies have not generally yielded satisfactory 
results. As Kitchen noted in his review of the basic work by Harris & 
Wente, their analyses offers an over-abundance of comparatively very 
young kings. 1 The cases of Thutmose III and Ramesses II are partic- 
ularly revealing. The estimates provided by Wente and Harris for the 
ages of the unequivocally identified mummies of these kings (40 and 
55 years, respectively) are glaringly at odds with the well-documented 

Harris & Wente, Atlas. 


long reigns of both rulers (53 and 66 years). "Something somewhere 
is badly wrong", as Kitchen remarked. 2 In other cases, as with Amenhotep 
III, the identification of the mummy itself is not beyond doubt, which 
adds to the uncertainty. 

The totals which the Manethonian sources cite include part of Dyn. 
19 and thus provide only a limited control. For the period from c Ahmose 
to Merneptah, Eusebius and the Old Chronicle give 348 years while 
Africanus has 263 years The tally resulting from the following list 
amounts to a minimum of 307 years and a maximum of 335 years for 
the same time span, effectively excluding Africanus' total. 

Dyn. 18 

Ahmose No data relating to the accession date is currently available. 
Most royal monuments are undated. Year 17 is cited on a stela from 
the foundations of the third pylon at Karnak (Abdul Q_ader Muhammad, 
ASAE 59, 1966, 148-149, pis. IV-V); year 18 is recorded on the stela 
Hanover 1935. 200. 209 (Im ^eichen des Mondes, Exhibition catalogue, 
Munich 1999, No. 4; A. Klug, Konigliche Stelen in der J?eit von Ahmose bis 
Amenophis III, Brussels 2002, 49-51, reviving doubts about its authen- 
ticity), and year 22 in the Turah quarries (Urk. IV 25,7). 

Josephus gives 'Ahmose a reign of 25 years, 4 months; Eusebius 
rounds this off to 25 years. (The figure is missing in Africanus.) The 
autobiography of an anonymous Viceroy of Nubia covers the period 
from 'Ahmose to Thutmose II (Urk. IV 39—41), and thus a maximum 
of 33 (21 + 12) years between the two rulers. For Ahmose, the min- 
imum reign length should be 21 years and the maximum 25 full years. 
His mummy (CG 61057) was estimated to be that of a man 25-30 
years old (Harris & Wente, Atlas, 202), but a general uncertainty pro- 
hibits drawing any conclusions. 

The precise date of the defeat of the Hyksos and thus the end of 
Dyn. 15 must lie in Ahmose's second decade. "Year 11" in the Rhind 
Mathematical Papyrus should probably be assigned to the last Hyksos 
ruler Khamudi (A.-F. el-Sabbahy, GM 133, 1993, 97-99, cf. above 
Schneider, Chapter II. 7), but Kitchen still maintains that the date 

K. A. Kitchen, JNES 44 (1985), 235-237. 


belongs to c Ahmose (in SCIEM, 2000, 45). Since P. Rhind refers to 
the conquests of Heliopolis and Sile, the capture of Avaris could fol- 
low somewhat later. 

Amenhotep I For the accession, W. Helck (in: Fs S. Schott, Wiesbaden 
1968, 71-72) assumed that the festival dates and the months in Josephus 
indicated 29-30/1/ Akhet, as opposed to D. B. Redford (JJVES 25, 
1966, 115-116) who used the same festival dates to argue for 11/111/ 
Shemu while Krauss (Sothis, 115) considers the actual date of P. Ebers, 
9/111/ Shemu, to be the accession date, as do U. Luft (GM 92, 1986, 
69-77) and Beckerath (Chronologie MR, HO). 3 G. Vittmann believes the 
title "Royal Mother" borne by c Ahmes Nefertari in the Turah inscrip- 
tion of year 22 (Urk. IV 25,4) supports a coregency; but he also stresses 
its chronological irrelevance, since Amenhotep I counted his regnal 
years from the death of his father ("Was there a coregency of Ahmose 
with Amenophis I?", JEA 60, 1974, 250-251). 

Dated monuments belong to year 7 (graffito of the Viceroy Tury in 
Semna: J. H. Breasted, AJSL 25, 1908, 108), year 8 (Uronarti: Urk. IV 
78,8; stela from Qasr Ibrim: J. Plumley, JEA 50, 1964, 4 with pi. 1,3), 
year 9 (rock inscription in Semna: F. Hintze, ^AS 111, 1984, 137—138 
and Hintze & Reineke, Felsinschriften, No. 512), and 1/I/Shemu of year 
10 (Kares Stela, CG 34003: Urk. IV 45,9). Thereafter there is only a 
graffito at Saqqara dated 19/IV/ Akhet of year 20 (Step Pyramid I, 79). 

Josephus assigns 'Ahmose 20 years 7 months; the other Manethonian 
sources round this up to 21 years, which accords well with the 2 1 years 
that the "astronomer" Amenemhat spent (jrj) under Amenhotep I 
(L. Borchardt, Altagyptische Ratines sung, Berlin & Leipzig 1920, pi. 18). 
Wente & Van Siclen argue that this refers only to sole rule and add 
6 years coregency with c Ahmose, but even this would not produce the 
30 years necessary for a "real" ^-festival which they presume ("Chron- 
ology", 225). The accession date of Thutmose I means fixes the death 
date of c Ahmose on 20/111/ Peret. 

Thutmose I The accession on 21/111/ Peret is certain (Urk. IV 81,4). 
Further dates are 15/11/ Akhet year 2 (Tombos: Urk. IV 82,9), 22/1/ 
Shemu year 3 return from Nubia (Urk. IV 88,11; 89,6/16), year 4 on 

On the problems of the Ebers date, see below Chapter III. 10. 


a naos from Giza (Urk. IV 91,9; perhaps posthumous or modern: 
R. Krauss, A & L 3, 1992, 86), and year 8/9 on a block from Karnak 
(A. Mariette, Karnak, Leipzig 1875, pi. 32. f), with doubtful attribution 
(see Wente & Van Siclen, "Chronology", 225-226, and R. Krauss, A 
& L 3, 1992, 86-87 with fig. 3). Perhaps the "11 years" on the stela 
of Nebwawy {Urk. IV 208,16) also refer to him, although D. B. Redford 
(JNES 25, 1966, 118-119) prefers Thutmose II. 

With an estimated age of 18-22 years, the mummy (CG 61065) seems 
to be extremely young (Harris & Wente, Atlas, 202) for a reign of 12 
years 9 months, according to Josephus. 

Thutmose II Following Gardiner, Redford accepts 8/11/ Akhet (JNES 

25, 1966, 117) as the accession date, whereas Beckerath (Chronologie NR, 
117) assumes that it took place in III/IV Akhet, on account of the 
months assigned to his predecessor by Josephus. The only certain date 
is 9/11/ Akhet of year 1 (Aswan: Urk. IV 137, 9). A short reign is sup- 
ported by the small number of scarabs (105 examples collected by 
B. Jaeger for Thutmoses II as opposed to 292 for Thutmoses I) 4 and 
gaps in the succession of officials, e.g. the stewards of Karnak, as well 
as the lack of a mortuary temple (signalled by L. Gabolde, "La chronolo- 
gie du regne de Thoutmosis II", SAK 14, 1987, 61-81). However, 
Beckerath (SAK 17, 1990, 65-74; Chronologie NR, 121), W. Barta [JEOL 

26, 1980, 33-34) and Wente & Van Siclen, "Chronologie", 226-227) 
have argued for a 13-year reign on the bases of the age of his chil- 
dren at the end of the reign, 5 the Ebers date, and the ^-festival of 
Hatshepsut, which is capable of different interpretations. 

The "Year 18" which often appears in this context does not belong 
to Thutmose II but rather to Amenhotep II, according to L. Gabolde 
(SAK 14, 1987, 61—81), or to Maatkare Hatshepsut, according to Beckerath 
(SAK 17, 1990, 66, cf R. Krauss, A & L 3, 1992, 86 with n. 3). The 
identification of the mummy (CG 6 1 066) is uncertain and thus its age 
estimate of 25—30 years is not relevant. Manetho assigned "Chebron", 
who is generally identified as Thutmose II, 13 years, but this could 
easily be an extension of 10 years; 3 years fit the sources, but they are 

4 Cf. A&L 3 (1992), 88. 

5 On the problem of the expression "in the nest", cf. R. Krauss, OL£ 90 (1995). 


no more than a possibility. At present, we can assume perhaps 2 to 4 
years for the reign of this king. 

Hatschepsut When Hatshepsut seized power continues to be debated. 
The oracle of 29/11/ Peret year 2, initially postulated by Schott, is not 
sufficiendy explicit (cf Ch. Cannuyer, in: Studies Lichtheim I, 109-115). 
The queen counted her own years from her coregent's accession, so 
that Thutmose Ill's accession date is taken for hers, and confirmed 
through Urk. IV 367,3—5 on the creation of her obelisks in 7 months, 
from 1/11/ Peret year 15 to 30/IV/ Shemu year 16. 

The first certain date is year 9 for the Punt expedition [Urk. IV 
349,10). 12/111/ Peret year 12 in West Tangur (Hintze & Reineke, 
Felsinschriften, 172, No. 562) is linked to both kings, as is year 16 in 
Wadi Maghara (Urk. IV 393,15), as well as an inscription of the over- 
seer of works Nakht in year 20 6 and a graffito at Saqqara of 2/111/ 
Peret year 20 (Step Pyramid I, 80, F). An inscription of 30/1/ Akhet 
year 17 in Karnak (Urk. IV 376,13) names Hatshepsut alone, and 
another of year 20 in Nubia (Urk. IV 1375,3) only Thutmose III. An 
ostracon of IV Peret year 20 from Deir el-Bahri should probably be 
assigned to Hatshepsut (W. C. Hayes, JEA 46, 1960, 38). 

Thutmose III His accession occurred with certainty on 4/1/ Shemu 
(Urk. IV 180, 15—16). The earliest document thereafter is a graffito at 
Saqqara (Step Pyramid I, 80, D) which mentions a royal sojourn at 
Thebes for 5/IV/ Akhet of year 1. His sole rule could have begun on 
10/11/ Peret year 22, recorded on a stela from Armant (Urk. IV 1244,14). 
Josephus gives Hatshepsut a reign of 2 1 years 9 months, which would 
cover her joint reign with Thutmose III. According to Urk. IV 895,16—17, 
Thutmose III died on 30/111/ Peret of his year 54. Before his sole 
rule, we have precise dates for a renewal of offerings in the Semna 
Temple on 8/11/ Shemu year 2 (Urk. IV 193,13), for the inauguration 
of the Vezier Useramun on 1/1/ Akhet year 5 (Urk. IV 1384,3), a 
mention of 26/11/ Akhet year 7 in the Akhmenu at Karnak (Urk. IV 
1256,8), a renewal of offerings in Karnak on 27/1/ Shemu year 15 
(Urk. IV 172,15); year 15 (without a royal name) is mentioned in the 
autobiography of Iamunedjeh (Urk. 940,5), year 16 in a rock inscrip- 

6 Urk. IV 1377,3. 


don at Abusir (Hintze & Reineke, Felsinschriften, 38 No. 64), and 16/IV/ 
Shemu year 18 at Shalfak (ibidem, 90 No. 365). 

The stela of Senimes bears the date 25/111/ Peret year 21 (Urk. IV 
1066,10). 25/IV/ Peret year 22 appears in the annals as the begin- 
ning of the first campaign (Urk. IV 647,12); the same year 22 also 
occurs in a reference to the restoration of a statue (Urk. IV 605,17). 
Dates from the Megiddo campaign include year 23, 4/1 Shemu for 
arrival at Gaza (Urk. IV 648,9) and 16/1/ Shemu for the council of 
war at Yehem (IV 649, 3ff). Year 23 is also documented from the tem- 
ple in Wadi Haifa (IV 806,11), and the Akhmenu at Karnak records 
2/1/ Shemu of a "year after 23" (IV 1252, 11, cf. A. H. Gardiner, 
JEA 38, 1952, 9). On 30/11/ Peret year 24 a foundation ceremony in 
Karnak is mentioned (Urk. IV 836,2), and for year 24 also a list of 
tributes (IV 671,6). Year 25 occurs in the Botanical Garden in Karnak 
(IV 777,2) and on a stela at Serabit el-Khadim, Sinai (IV 886,5), year 
27 on the statue of Sabastet (IV 1369,4), year 28 in the tomb TT 82 
(IV 1043,15), year 29 in the annals for the fifth campaign (IV 685,3), 
and year 30 for the sixth campaign (IV 689,3). 

3/1/ Shemu year 31 is mentioned in the list of booty for the sev- 
enth campaign (Urk. IV 690,14). An inscription of Sennefer at el-Bersheh 
(Urk. IV 597,12) cites 12/IV/ Shemu year 33 in relation to the sed- 
festival (a parallel inscription in Krakow gives 4/IV/ Shemu); year 33 
is documented from the annals for the eighth campaign (Urk. IV 696,15), 
year 34 for the ninth campaign (Urk. IV 703,16), and year 35 for the 
tenth campaign (Urk. IV 709,15); year [38] can be restored for the 
thirteenth campaign (Urk. IV 716,12); year 39 for the fourteenth (Urk. 
IV 721,9) is also known from a graffito at Saqqara (13/111/?: Step 
Pyramid I, 80, E), and year 40 from a list of tribute (Urk. IV 668,4). 

22?/I/ Akhet year 42 is documented with the restoration of a statue 
in Karnak (Urk. IV 606,6), year 42 at the end of the annals (IV 734,14); 
2/11/ Peret year 43 is known from an ostracon from the west bank of 
Thebes (IV 1374,8), and year 45 from the dedication for a Mnevis- 
bull (IV 1373,2); 10/111/ Akhet year 47 is the date of the Gebel Barkal 
stela (IV 1228,6), year 47 of the Mnevis-bull stela Cairo JdE 65830 
(M. Moursi, SAK 14, 1987, 233-235), and of the stela Berlin 1634 con- 
cerning an enclosure wall in the temple of Heliopolis (Urk. IV 832,12); 
22/1/ Shemu year 50 is from a rock inscription at Sehel (Urk. IV 
814,10; cf. also J. Leclant & G. Clerc, Or 61, 1992, 299). 14/II/Shemu 
year 51 is the date of a rock inscription at Ellesia (Urk. IV 811,10), 
and 23/111/ Peret year 53 is found on a scribal palette in Hanover 
(W. Helck, MDAIK48, 1992, 41-44). 


The reign length which Josephus associated with Amenhotep II is 
30 years 10 months (reduced by ten years by Theophilus), and this 
could be identified as the sole rule of Thutmose III, if one assumes a 
coregency with his son, but might actually be Amenhotep III, if his 
reign length be reduced by ten years. 

Amenhotep II The accession date (or the date when he was named core- 
gent) was 1/IV/ Akhet (Urk. 1343,10). According to Bierbrier (Or 49, 
1980, 108), who argues against a coregency, this is in error for 1/IV/ 
Peret. Many authors side with Redford (JEA 51, 1965, 107-122) and 
Parker (in Studies in Honor of John A. Wilson, Chicago 1969, 75—82), in 
favor of a coregency lasting 2 years 4 months. Despite P. Der Manuelian's 
lengthy discussion in his Studies in the Reign of Amenophis II (Hildesheim 
1987), 19-40, the matter remains unresolved, cf. R. Krauss, OL£ 90, 
1995, 241-242. 

The earliest date is 15/111/ Shemu year 3 (Amada stela: Urk. IV 
1289, 1) which, in the event of a coregency, might mark the begin- 
ning of sole rule. To year 4 belong an inscription from the Turah 
quarries (IV 1448,4), and an addition to the duplicate of the Amada 
stela in Cairo (IV 1299,2), as well as the 10/1/ Peret graffito at Saqqara 
(Step Pyramid I, 80, G). The years 5, 6, 7 and 9 are documented by 
several dates on stelae from Memphis and Karnak (Urk. IV 1301—1314), 
but there follows a gap until at least year 17 (uncertain: L. Habachi, 
Kemi 18, 1968, 55 with fig. 5 on the graffito of Nakht in Aswan) or 
even until 1/IV/ Akhet year 23, the date of the stela of the Viceroy 
Usersatet from Semna, in Boston (Urk. IV 1343,10). 

A jar label from the mortuary temple names year 26 (Urk. IV 
1365,18-20), which suits the 25 years 10 months of Josephus, but unfor- 
tunately, it is not certain that this figure belongs to Amenhotep II, 
although it makes no sense for Thutmose III either. No available infor- 
mation suggests a longer reign. Only the upper limit for the estimated 
age of the mummy (CG 61069; Harris & Wente, Atlas, 202) at 35-45 
years can be reconciled with the historical data. 

Thutmose IV There is no basis for determining the date of his acces- 
sion, other than Josephus' figure of 8 months which suggests the end 
of II or beginning of III Akhet. On this assumption the date of the 
sphinx stela, 19/111/ Akhet year 1 (Urk. 1540,2) is plausibly the earli- 
est of the reign. Somewhat later in the first year would be 7/11/ Shemu 
from a stela in Luxor (El Sayed Higazy & B. M. Bryan, "A New Stela 
of Thutmose IV from the Luxor Temple", VA 2, 1986, 93-100). Bryan 


provides a list of documented years in the reign (The Reign of Thutmose 
IV, Baltimore & London 1991, 5-9. 

An inscription on Sinai (Urk IV 1634,9) is dated to year 4, others 
to years 5 and 7 (IV 1564). A stela on the art market, probably from 
Medamud, names 10/1/ Shemu year 5 (R. G. Bigler & B. Geiger, 
"Eine Schenkungsstele Thutmosis' IV", £4'S* 121, 1994, 11-17), two 
Theban tomb inscriptions (ID Text III, 273 and Urk. IV 1618,5) year 
6; year 7 (to be corrected to 8) appears on a stela from the island of 
Konosso (IV 1555-1556). The highest date is 2/111/ Peret year 8 on 
the Konosso stela (IV 1545,6), when the king is informed about a 
Nubian rebellion. 

Josephus gives "Thmosis" 9 years 8 months. The other Manethonian 
sources round down to 9 years and the Book of Sothis has an inflated 
39 years. Years 9-10, like years 2 and 3, are not documented in con- 
temporaneous sources. This seems acceptable, whereas a longer reign 
results in difficulties, and certainly the extreme lengthening proposed 
by Wente & Van Siclen on the basis of their ideas about sed-festivals 
is precluded. The age at death of his mummy (CG 61073) was esti- 
mated at 30-40 years (Harris & Wente, Atlas, 202; cf. Bryan, Reign, 
9—13). According to the inscription on the Lateran obelisk (Urk. IV 
1550,5), it lay "on its side for 35 years" before being erected, which 
would include the entire reign of Amenhotep II and an unknown num- 
ber of the years of Thutmose III and IV. 

Amenhotep III Based on the date of the first ^-festival, Helck (Manetho, 
67) argued for an accession on 3/111/ Shemu. An inscription from III 
Shemu year 1 was found in el-Bersheh (Urk. IV 1677—1678); in year 
2 the quarries at Turah were opened (Urk. IV 1681,2) and the com- 
memorative scarab of the wild bull hunt was issued (Urk. IV 1739,1). 
A graffito in the royal tomb WV 22, at the passage into the antecham- 
ber, is dated to 7/111/ Akhet year 3 (J. Kondo, in: R. H. Wilkinson 
(ed.), Valley of the Sun Kings, Tucson 1995, 29-30, fig. 3); a wine jar 
label from Deir el-Medina also mentions year 3 (Valbelle, Ouvriers, 23). 
Several dates in year 5 are linked to the Nubian campaign: 24/11/ 
Akhet on the island of Sai (Urk. IV 1959,11), 2/111/ Akhet near Aswan 
(Urk. IV 1665,15), I Shemu at Buhen (Urk. IV 1758,17) and a stela at 
Konosso on the return from Kush (Urk. IV 1662,7). The lion-hunt 
commemorative scarab dates to year 10 (Urk. IV 1740,12), as does the 
scarab commemorating the wedding with Gilukhepa (Urk. IV 1738,6), 
and a graffito of 13/111/ Shemu at Saqqara (Step Pyramid I, 81, H). 


The second decade of the reign is poorly documented. The series of 
commemorative scarabs ends with that relating to the excavation of a 
lake for Queen Teye, dated 1—16/111/ Akhet in year 11 (Urk. IV 
1737,8). Labels on wine jars from Deir el-Medina document years 14, 
17 and 19 (Valbelle, Ouvriers, 23); a papyrus fragment in Turin men- 
tions IV Peret year 14 (GLR II 310, XI). For II Akhet year 20 the 
statue of Nebunefer in Brussels records a royal visit to Memphis (Urk. 
IV 1885,10), and 2/1/ Peret (without royal name) is documented in 
an historical inscription from the mortuary temple of Amenhotep-son- 
of-Hapu (A. Varille, Inscriptions concernant I'architecte Amenhotep fils de Hapou, 
Cairo 1968, 96-97). 

The third decade is likewise sparsely documented. A jar-label from 
Amarna of year 21 should belong to this reign (JEA 67, 1981, 2), as 
should the dates of 20/111/ Shemu year 27 of P. Berlin 9784 (A. H. 
Gardiner, £4$ 43, 1906, 28-35) and O. Cairo CG 25242 of 20/IV/ 
Akhet year 29. Many dates are associated with the first .y^-festival in 
year 30, beginning on 27/11/ Shemu (Kheruef: Urk. IV 1867,2) and 
extending through III Shemu (Urk. IV 1869,2; 1837,9); these are sup- 
plemented with numerous labels from Malqata, some from year 29, 
but generally mentioning year 30 and the first .^-festival (W. C. Hayes, 
JNES 10, 1951, 35-56, 82-112, 156-183, 231-242, and additional 
labels in M. A. Leahy, Excavations at Malkata and the Birket Habu 1971—1974. 
The Inscriptions, Warminster 1978). 

Further deliveries at Malqata are dated to year 31, along with the 
hieratic stela BM 138 with the decree for Amenhotep-son-of-Hapu 
(6/IV/ Akhet: Varille, Inscriptions, 67-85; D. Wildung, Imhotep und 
Amenhotep, Munich & Berlin 1977, § 182). A contract from Gurob is 
dated 5/1/ Akhet year 33 (P. Gurob 11,1: A. H. Gardiner, £AS 43, 
1906, 35-37), and the second ^-festival of year 34 is documented with 
numerous labels from Malqata (Hayes, in JNES 10, 1951). 1/1/ Shemu 
year 35 is the date of two stelae at Silsileh (Urk. IV 1920,3; 1678,8); 
the third .^-festival of year 36 is documented in the tomb of Kheruef 
(Urk. IV 1860,2-7), and 9/11/ Peret of year 36 on a Sinai stela (Urk. 
IV 1891,4). Labels from Molkata mention year 37 for the third sed- 
festival (Urk. IV 1954,12), stretching through 1/111/ Shemu of year 38 
(Hayes, JNES 10, 1951, fig. 11, no. 142). So at least 37, and perhaps 
38 full years should be assumed for Amenhotep III; Manetho's copyists 
give 38 years 7 months, as well as 36 and 37 years (Helck, Manetho, 67). 

A long coregency with Akhenaten has been posited time and again 
since 1899 (Petrie, A History of Egypt II, London, 208), but it has been 


challenged by W. Helck (MIO 2, 1954, 196-209), and many others, 
most recently by Gabolde, Akhenaton, 62-98, who discusses most of the 
arguments. Gardiner's remark years ago (Egypt of the Pharaohs, Oxford 
1961, 213) still stands: "the much canvassed co-regency must be an 

Amenhotep IV/Akhenaten Analysis of the sequence of dates on the later 
boundary stelae at Amarna, Akhenaten's newly founded capital, places 
the accession in the time span 1/1 Peret — 8/1 Peret (see W.J. Murnane, 
"On the Accession Date of Akhenaten", in Studies in Honor of George R. 
Hughes, Chicago 1976, 163—167; Murnane & Van Siclen, Stelae, 155; 
Beckerath, Or 63, 1994, 123; Gabolde, Akhenaton, 14-16). 

One of the earliest dates of the reign is the hieratic docket on the 
Amarna letter EA 27, from 5?/I/ Peret year 2 (Urk. IV 1995,16, cf 
W. Fritz, SAK 18, 1991, 207-214). 27/?/Peret year 2 and epagome- 
nal day 4 of year 3 are mentioned in P. Berlin 9784 (Murnane, Texts, 
44), 7/11/ Akhet year 4 in P. Berlin 9785 (Murnane, Texts, 46), 11/111/ 
Akhet year 4 with the expedition of the High Priest May in the Wadi 
Hammamat (Murnane, Texts, 68), 19/111/ Peret year 5 in the letter of 
Ipi to the king (Murnane, Texts, 51). The earlier series of boundary 
stelae is dated 13/IV/ Peret year 5, the later series exactly a year later, 
with repetition of the oath on 8/1/ Peret year 8 and a colophon of 
30/IV/ Akhet year 8 (Murnane & Van Siclen, Stelae, 73—86). There is 
a reference to "Year 9 of the heretic" in a letter from the reign of 
Ramesses II (P. Berlin 3040: KRI III 5). 

The later years are less well documented in monumental inscriptions. 
8/11/ Peret year 12 is mentioned in the tribute scenes in the tombs of 
Meryre and Huya (Urk. 2003,5; 2006,11); 20/111?/ Akhet year 12 is 
the date of the Nubian campaign on a stela in Buhen (Murnane, Texts, 
101-102); 2?/II/ Peret year 14 of a graffito at Saqqara (J. Malek, DE 
32, 1995, 105—106). Wine jar labels from the site of Amarna docu- 
ment Akhenaten's regnal years 4 through 17 (Hornung, Untersuchungen, 
90-91; Krauss, MDOG 129, 1997, 227-229), so that he could have 
died at the earliest during the sealing of the wine jars in year 17 (II 
Akhet, corresponding to August 22 to September 20 [Julian] in 1336 
BC) or just before the sealing in year 1 8 at the latest. Akhenaten does 
not seem to have been preserved in the Manethonian tradition. The 
conventional identification of QPOZ (Horos) with Akhenaten is problematic. 


Smenkhkare' and c Ankh(et)kheperure c It is now certain that not only a man 
c Ankhkheprure c , but also a woman c Ankhetkheprure c ruled between 
Akhenaten and Tut'ankhamun. The king is first known as 'Ankhkhe- 
prure c (throne name) Smenkhkare' dsr-hprw (personal name), later as 
'Ankhkheprure' mrjj /NeferkheprureVWaenreVand Nefernefruaton mrjj 
Waenre c . The two names of the queen, i.e. c Ankhetkheprure c mrjj/ 
NeferkheprureVWaenreV and Nefernefruaton >ht n hj.s, are nearly the 
same as the king's later set of names and epitheta. The "funerary" epi- 
theton Iht n hj.s (beneficial for her husband) is hers alone and indicates 
that she succeeded her husband c Ankhkheprure c . Her identity remains 
problematic; Kiya, Nefertiti, and Merytaton have been proposed. Items 
of her funerary equipment were adapted for Tut'ankhamun (M. Gabolde, 
Egypte Afrique & Orient 33, 2004, 19-26). Josephus lists three rulers 
named AXENXEPPZZ < c Ankh(et)kheprure c , i.e. two male rulers (one of 
which might be due to a corruption in the text) and a female, described 
as a king's daughter. In its transmitted form the Manethonian tradi- 
tion ascribes 1 2 years and some months to either of the kings named 
Akhenkherres. Possibly the figures reflect an original 2 years and some 
months. The mummy of King c Ankhkheprure c is identifiable as the 
occupant of KV 55; his age at death was 18—22 years (W.J. Murnane, 
OLZ 96, 2001, 22). 

The wine jar labels from Amarna attest sixteen successive wine vin- 
tages during the occupation of the site, 13 corresponding to years 5 
through 17 of Akhenaten, whereas 3 vintages correspond to regnal 
years of his successors. The vintage of Akhenaten's year 4 occurred in 
the year before the foundation of the city; wine of year 4 was con- 
sumed at the site before the vintage of year 5 became available. Thus 
altogether seventeen successive vintages are attested at Amarna. 

Up to year 1 3 of Akhenaten the chief vintners held the title hrj kSmw. 
The title hrj b>h is attested from year 13 through 17 and its use con- 
tinued in year 1 and 2 of King c Ankhkheprure c . The last vintage that 
is documented at Amarna dates to a regnal year 1; in that year the 
vintner's title hrj kHmw was reintroduced and continued to be used as 
wine jar labels in the tomb of Tut'ankhamun show (see below). Regnal 
year 1 of the last vintage at Amarna could belong either to the queen 
'Ankhetkheprure' or to Tut'ankhamun. The only inscription from this 
period that is dated by a regnal year with a royal name is the graffito 
in the Theban tomb of Pairy: 10/III/Akhet year 3 of c Ankhkheprure c 
mrjj /// Nefernefruaton mrjj /// (Urk. IV 2024, 14). The date seems 
to relate to the king, but the queen is not excluded (Gabolde, Akhenaton, 


161—162, 184). If the graffito relates to her, then she continued the 
year count of her husband and the last vintage that is attested at 
Amarna dates to year 1 of Tut'ankhamun. If the graffito relates to the 
king, then the queen started a regnal year count of her own; year 1 
of the last vintage belongs to her and there would be no dated mate- 
rial of Tut'ankhamun at Amarna. 

A regnal year 3 is also attested at Amarna in the labels on vessels 
for various commodities. Year 3 continues year 1 and 2 of King 
c Ankhkheprure c as labels of year 2 and 3 belonging to a single deliv- 
ery of olive oil prove (Hornung, Untersuchungen, 88—89). There are only 
3 wine jar labels of year 3 which cannot represent a complete vintage, 
because the yearly mean number of wine jar labels is 50 to 60. The 
disproportion is explicable if the change from regnal year 2 to 3 occurred 
during the sealing of the wine jars. Thus King c Ankhkheprure c would 
have counted his reign from a day in ca. II Akhet (Krauss, MDOG 
129, 1997, 238), which may have coincided with the occurrence of 
Akhenaten's death. 

Tut'ankhamun There are no plausible proposals for his accession date, 
nor by dated inscriptions for his first three years. A graffito at Saqqara 
is dated 2/IV/ Shemu year 4 (Step Pyramid I, 78); year 4 is also attested 
on a donation stela (W. Kaiser et al., Agyptisches Museum Berlin [1967] 
no. 776); a label on a textile from his tomb cites year 6 (Urk. IV 2062,4); 
the decree for the Overseer of the Treasury Maya is dated 22/111/ 
Peret year 8 (Stela Liverpool E. 583: A. A. M. A. Amer, "Tutankhamun's 
Decree for the Chief Treasurer Maya", RdE 36, 1985, 17-20). Wine 
jar labels from the tomb document years 4, 5, 9 and 10 (J. Cerny, 
Hieratic Inscriptions from the Tomb of Tut'ankhamun, Oxford 1965, 1-3), but 
the last could also have been Akhenaten's (R. Krauss, OLJ? 90, 1995, 
245-246, and P. Tallet, BIFAO 96, 1996, 369-383). 

Analysis of botanical remains from the funeral show that the burial 
took place in March/April, and thus the death in January, III or IV 
Peret (R. Krauss, "Nochmals die Bestattungszeit Tutanchamuns", SAK 
23, 1996, 227-254). The age of the mummy was estimated at ca. 18 
years (F. Filce Leek, The Human Remains from the Tomb of Tut'ankhamun, 
Oxford 1972). 

Aya The death of Tut'ankhamun in III or IV Peret should date the 
accession of Aya. A wine jar label from Deir el-Medina mentions Aya's 
mortuary temple and a year 2 (KRI VII 65,9), but the date could belong 


to Haremhab's reign. A donation stela of the king is dated to 1/111/ 
Shemu year 3 (Urk. IV 2109,8), and two others bear the date 1 /TV/ 
Akhet year 4 (Urk. IV 2110,13), the highest date known, which accords 
well with Josephus' 4 years 1 month. 

Haremhab The accession may have been in II Akhet, provided it was 
related to the king's participation in the Opet Festival (Hornung, 
Untersuchungen, 38-39). But the change of year from 26 to 27 in O. 
IFAO 1254 leads Krauss to conclude that the accession may be placed 
in the interval between 28/IV/ Peret and 13/1/ Shemu (DE 30, 1994, 
73-85). However, it is not certain that these dates should be assigned 
to Haremhab. 

Festival foundations of 22/IV/ Akhet year 1 are known from Karnak 
(Urk. IV 2132,4); the rewards of Neferhotep in TT 50 (Urk. IV 2177,6) 
date to year 3; 1/1 /Akhet year 6 is attested on a stela from the mor- 
tuary temple of Amenhotep III (G. Haeny, BABA 11, 1981, 65-70); 
year 7 is mentioned in O. BM 5624 (IV 2162,10); and Maya's graffito in 
the tomb of Thutmose F/ is dated in III Akhet year 8 (Urk. IV 2170,15). 
Helck compiled a number of additional dates from years 1 to 13 (A 
& L 3, 1992, 64), and the highest certain date at present is III Akhet 
year 13, in a wine jar label from Saqqara (Martin, JEA 65, 1979, 15; 
cf Murnane, Texts, 107 C 2). Hari (Horemhab, 392) wanted to assign a 
year 20(?) in El Kab (Capart, ASAE 37, 1937, 10) to Haremhab. Krauss 
argues that O. IFAO 1254 supports a longer reign (DE 30, 1994, 
73-85), and a longer reign is also favored by Beckerath (SAK 6, 1978, 
43-49) and van Dijk (GM 148, 1995, 29-34: 15-17 years). 

The interpretation of the graffito in the royal mortuary temple refer- 
ring to the "day of entering of King Haremhab", dated to 9/1/ Shemu 
year 27, is contentious. Beckerath (Tanis und Theben, Gliickstadt 1951, 
104) argued that this was the date of death (better: burial); the use of 
the expression jrjt c kw n with the meaning "burial" is occurs in the con- 
temporaneous TT 50 (Hari, La tombe thebaine du pere divin Neferhotep, 
Geneva 1985, pi. X), and c kw is also used for the burial of the Apis- 
mothers (Smith, in Lloyd, ed., Studies J. G. Griffiths, London 1992, 
203-204). Thus it seems legitimate to understand this date as a refer- 
ence to the burial of Haremhab, which suggests that he died at the 
end of II Peret. O. IFAO 1254 also points to year 27, and this in turn 
is compatible, without any emendation, with the year 58 or 59 of the 
Mes inscription, including the reigns of all the proscribed kings of the 
Amarna Period. 


Dyn. 19 

With the exception of the relationship between Sety II and Amenmesses, 
the royal succession of this dynasty is not a matter of debate, and, 
excluding some minor details, the lengths of the various reigns are like- 
wise relatively certain. Helck provided a list of the confirmed regnal 
years from Ramesses I to year 10 of Ramesses II (A & L 3, 1992, 64), 
and Wente & Van Siclen compiled another covering the time span 
from year 32 of Ramesses II through Ramesses X ("Chronology", 
251-261). Kitchen's index (KRI VIII 70-84), supplemented by Helck, 
Ostraka, is more comprehensive for Ramesses I through Ramesses XL 
However, as many dates are not linked to a royal name, assignment 
to particular reigns are subject to change. Demaree has supplied an 
overview of accession dates from Sety I through Ramesses XI (GM 
137, 1993, 52). 

The surviving portion of Manetho's list for the later phase of the 
dynasty is problematic since only Twosre's ("Thoeris") reign is correct 
at 7 years, including the coregency with Siptah and her sole rule. 
Africanus provides a total of 209 years, and Eusebius 194, but both 
are too high as the actual sum cannot be more than 104 years. 

Ramesses I A possible date for his accession can be deduced using Sety 
I's accession date in combination with the months indicated by Josephus 
(neither of which is certain, however), resulting in III, or possibly IV 
Peret. On this basis, the only certain date of the reign, 20/11/ Peret 
year 2 on the stela Louvre C 57 from Buhen (KRI I 2,9), would lie at 
the end of the second year of the reign. The king may have died at 
the beginning of his third year, which would allow for two full years. 
(For dates from his reign see also Krauss, DE 30, 1994, 83, n. 13.) 
Josephus assigns him 1 year 4 months which Africanus rounds down 
to one year, allowing for an estimate of one or, alternatively, two years. 

Sety I Utilizing the date of a later processional festival celebrating Sety 
I (KRIVL 249,7), Helck (CdE 41, 1966, 233-234; SAK 17, 1990, 207-208) 
and Krauss (OL-Z 90, 1995, 246; SAK 24, 1997, 168) propose 24/111/ 
Shemu (cf. KRI VIII 70) for the accession. Basing his analysis on palace 
accounts in Memphis, Murnane identified a period between 18/111 to 
23/IV/ Shemu, and favored the festival date of 2/IV/ Shemu (Serapis 
3, 1975-1976, 23-33). Using pRollin 1889 Redford {King-Lists, 113 with 
n. 46) chooses 23/IV/ Shemu, or the interval between 30/11/ Shemu 


and 2/1/ Akhet, which is compatible with the festival date. For Brand 
{The Monuments ofSetil, Leiden 2000, 302), 24/111/ Shemu is "the most 
likely solution", but the date cannot be considered certain. 

Analyzing the inscription of Bakenkhons, Bierbrier reckons the length 
of the reign at 15 years (JEA 58, 1972, 303), and Kitchen (JNES 39, 
1980, 170-171; High, Middle or Low I, 3, and III, 153-154) concurs. 
However, Jansen-Winkeln argues {JNES 52, 1993, 221-225) that the 
autobiographical details of the high priest's career do not allow for such 
extrapolations. Spalinger suspects that 10 years and a fraction is more 
appropriate, given the data from the campaigns (JARCE 16, 1979, 41 
n. 106). The abundance of sources for years 1-11 suggests that Sety I 
died in year 11; only year 10 is missing (KRI VIII 70), while 13/IV/ 
Shemu year 1 1 is known from Gebel Barkal (KRI I 75,8). The limited 
production of the Aswan quarries which were opened in year 9 (Brand, 
JARCE 34, 1997, 101-114) favors 11 years, as does the extended sum 
of 51 regnal years in Africanus (Josephus gives 59). After discussing 
the sources, Brand favors 11 full years, or perhaps 10 years (Monuments 
of Seti I, 305-309), and Kitchen now suggests 11-15 years (in SCIEM, 
2000, 42-43). 

Arguments for a coregency between Sety I and Ramesses II (Murnane, 
Ancient Egyptian Coregencies, Chicago 1977, 57-87, 183-185) have been 
criticized by Kitchen (JNES 39, 1980, 169—171), and Brand also comes 
to a negative conclusion (Monuments of Seti I, 312-332. Regardless, 
Ramesses II's regnal year count did not begin under Sety I. Clearly, 
the kings of the Ramesside period found a way for the designated suc- 
cessor to share in sovereignty other than using a titulary and separate 
dating system. 

Ramesses II Helck suggested 27/111/ Shemu for the accession of Ramesses 
II (AnBib 12, 1959, 118-120); Krauss (SAKS, 1977, 146-148), Valbelle 
(Ouvriers, 167) and Demaree (GM 137, 1993, 52) concur with this date 
which is that of a regularly documented festival. An alternative sugges- 
tion made by Larson and Wente & Van Siclen was countered by Helck 
(SAK 17, 1990, 205-207). After exhaustive discussion, both Beckerath 
(Chronologie NR, 68—70) and Brand (Monuments of Sety /, 302—305) agree 
with Helck and Krauss. 

The king certainly ruled for 66 full years (66 years 2 months accord- 
ing to Josephus), and every year of his reign is documented (KRI VIII 
70-73), with the highest date being 18/1/ Akhet of year 67 (RAD 
30,10), soon followed by 19/11/ Akhet of a year 1 (RAD 30,14). There 


is an indication that he died after 29/1/ Akhet (Valbelle, Owners, 176 
with n. 4). 

Merenptah O. Cairo CG 25504 shows that the accession occurred 
between 19/1 and 13/11/ Akhet. Peden ("A Note on the Accession 
Date of Merenptah", GM 140, 1994, 69) narrows this interval to 
3—13/11/ Akhet, on the basis of a Theban graffito from 2/11/ Akhet 
year 1 (but cf Krauss, OL? 90, 1995, 246-247). Other options are 
19/1/ Akhet (Demaree, GM 137, 1993, 52, opposed by Kitchen, BiOr 
60, 2003, 586) and 5-7/11/ Akhet (Beckerath, GM 191, 2002, 5-6). 

The highest date is TV/ Shemu year 9 (O. Gardiner 197: KRI IV 
159,5). The report of his death reached Deir el Medina on 16/?/ Peret 
(HO 64,1,1); presumably he died during year 10. 7 /TV/ Akhet year 
10 is mentioned in P. Sallier I, 3,4 (LEM 79), and KRI TV 160 also 
assigns two Theban graffiti of 7 and 13/11/ Akhet year 10, announc- 
ing the inundation, to the reign of Merenptah. The 19 years 6 months 
given by Josephus can thus be reduced by a decade. Sety II was his im- 
mediate successor, without any intervening rule by Amenmesses (Krauss, 
SAK24:, 1997, 174-177). 

Sety II Helck calculated that the accession took place between the end 
of I Peret and the beginning of III Peret (AnBib 12, 1959, 123). He is 
followed by Janssen (Varia, 101 with n. 21), whereas Krauss posits an 
interval of 29/11/ Peret through 3/111/ Peret (extended up to 6/111/ 
Peret by Beckerath, Chronologie MR, 71). For Demaree (GM 137, 1993, 
52), it is 29/11/ Peret, announced on 16/?/ Peret in HO 64,1,1. 
According to O. Cairo CG 25560 (KRI TV 302), the king was in Thebes 
on 10/11/ Akhet year 1. 

The king's death was reported in Thebes on 19/1/ Peret year 6 (O. 
Cairo CG 25515: KRI TV 322); O. CG 25516 from 17/1/ Peret is still 
dated to his reign (KRI TV 328). A graffito above the tomb of Twosre 
dates his burial to 11/111/ Peret in year 1 (of Siptah; Altenmuller, in 
Reeves, ed., After Tut'ankhamun, London & New York 1992, 148, fig. 19). 
The mummy (CG 61037) suggests that Sety II died young. 

Amenmesses Janssen (Varia, 99-109), Krauss (SAK 24, 1997, 161-181) 
and Beckerath (Chronologie JVR, 70-73) have recently discussed the rela- 
tionship between the reigns of Amenmesses and Sety I. We know from 
HO 64,1 that the Foreman Neferhotep was absent from work after the 
accession of Sety II; thus his murder by "the enemy" can only have 


taken place later if the enemy be identified with Amenmesses or, rather, 
his followers. The date of accession can be placed between 27/I/Shemu 
and 18/111/ Shemu (O. Cairo CG 25783 & 25784: KRI IV 227,6). He 
was in power in Thebes during years 3 and 4 (perhaps earlier in Nubia), 
which are unaccounted for in the dates of Sety II, but are followed by 
a great "clean-up" in Deir el-Medina by Sety II. Treating Amenmesses 
as a rival king provides the best explanation for the various phases in 
the decoration of Sety IPs tomb, as pointed out by Dodson ("The 
Decorative Phases of the Tomb of Sethos II and their Historical 
Implications", JEA 85, 1999, 131—142), and also for the interruption 
of work in the tomb of Twosre (Altenmuller, in After Tut'ankhamun, 
141-164, esp. 149, 159). The last date of the rival king is 29/111/ 
Shemu year 4 (O. Cairo CG 25784, 15 — the only ostracon from this 
year!). Assigning ostraca from Deir el-Medina dated in years 1 and 2 
to Amenmesses (so Helck, Ostraka, 97) cannot be justified. 

Siptah Helck's suggestion that O. Cairo CG 25521 allows the acces- 
sion to be placed between 28/IV/ Akhet and 3/1/ Peret (AnBib 12, 
1959, 123-124) presumes a scribal error, but is supported by the year 
change in P. Greg between 28/IV/ Akhet and 11/1/ Peret (Janssen, 
Varia, 116, and Beckerath, Chronologie NR, 74). Following Helck (Studies 
Kakosy, 270), Demaree sets the date as 2/1/ Peret (GM 137, 1993, 52). 
Given the report of the death of Sety II on 19/1/ Peret (see supra), 
the accession must have taken place early in I Peret. 

The execution of the "Chancellor" Bay was announced on 27/111/ 
Shemu year 5 (Grandet, BIFAO 100, 2000, 339-342); therefore he can- 
not have been buried on 22/IV/ Akhet in year 3, as Altenmuller sug- 
gested {SAK 23, 1996, 1-9, and GM 171, 1999, 13-18). This date 
(without year) is better linked to the burial of Siptah by the Vizier Hori 
(O. Cairo CG 25792: KRI IV 414-415). To year 6 belongs a graffito 
found in Buhen (KRI IV 365 [2]). Siptah died between 9 and 12/11/ 
Akhet and was buried on 22/IV/ Akhet year 7, according to Beckerath 
(Chronologie NR, 74, following Helck, in Studies Kakosy, 270). Alternatively, 
Krauss (OLZ 90, 1995, 247-248) and Schneider (%AS 130, 2003, 144, 
146) suggest year 6. 

Twosre Work in her tomb began in year 2 of Sety II (W. Helck, SAK 
17, 1990, 208-210), but was interrupted for a time (see Amenmesses). 
After her debut as sole ruler, the queen counted her years as a con- 
tinuation of the deceased Siptah's reign. 


A graffito in Deir el-Bahri mentions a "visit" of Amun to the mor- 
tuary temple of the reigning pharaoh Twosre on 28/11/ Shemu year 
7 (Marciniak, Inscriptions, 59—60, No. 3). A date 5/111/ Peret year 8 
(O. Deir el-Medina 594: KRI IV 407,16) is assigned by Krauss (SAK 
4, 1976, 191, and OLZ 90, 1995, 248 n. 29) to Merenptah's reign 
instead of hers. For IV [Peret?] of year 8 on O. Cairo 25293 see 
Altenmiiller, JEA 68, 1982, 1 14 who suggests that she died in I Shemu 
year 8, which would correspond to the reign of 7 years recorded by 
the Manethonian copyists. 

Dm. 20 

The ancient compilers who excerpted Manetho's history did not record 
any names for this dynasty, simply citing "12" (actually 10) kings who 
reigned for 135 (Africanus), 172 (Eusebius) or 178 years. Only Africanus' 
total could be correct, provided it follows directly on the sum down to 
Merneptah and so includes the last rulers of Dyn. 19. The actual length 
of the various reigns is both certain and precise, thanks to an abun- 
dance of dated documents. The only significant change resulting from 
recent scholarship is a shortening of the reign of Ramesses X. Parker 
assigned him 9 full years based on a presumed lunar date, whereas 
contemporaneous sources do not justify more than 3 years. Problems 
were also created by a hypothetical "interregnum" at the inception of 
the dynasty, but again, our sources preclude such a proposal; there is 
simply no space available. 

Demaree (GM 137, 1993, 49-52) and Helck (Ostrakd) have discussed 
the accession dates, which are in general clearly restricted. Documented 
regnal years are listed in KRI VIII. 

Sethnakhte His accession date is not known, but Beckerath proposes the 
beginning of II Shemu (Studies Kakosy, 63-67; Chronologie JVR, 75-76). 
Dates are only known from year 2, on 10/11 and 24/111/ Shemu, but 
possibly also (without month) from year 3 on Sinai (Beckerath, Studies 
Kakosy, 63-67). He probably died on 25/1/ Shemu, at the start of his 
year 3 (Altenmuller, GM 145, 1995, 29-36). If his year count subsumes 
the sole rule of Twosre, he will have ruled for only a year; however, 
the graffiti Deir el-Bahri 3 and 10 are linked to the Festival of the 
Valley, and imply an interval of 10 years between year 7 of Twosre 
and year 7 of Ramesses III (Krauss, Sothis, 139), so that Sethnakhte 
must be assigned some 3 years. 


Ramesses III His accession was on 26/1/ Shemu (Beckerath, Chronologie 
NR, 11 , with n. 431). Every year of the reign is documented, except 
for the first, down through year 32. His death occurred on 14P/III/ 
Shemu year 32. The burial equipment was transported into the Valley 
of the Kings on 4/1/ Akhet, and the burial will have taken place on 
24/1/ Akhet (O. Deir el-Medina 40: KRI VI 106, see Cerny, £45* 72, 
1936, 113, and Helck, Studies Kakosy, 269). According to information 
provided by O. Chicago 12073 (Allam, Ostraka, No. 40, 73-76) 18 years 
elapsed between year 1 7 of Ramesses III and year 3 of Ramesses IV. 

Ramesses IV His accession on 15/111/ Shemu is certain (O. Deir el- 
Medina 44); word reached the workers at Deir el-Medina on the fol- 
lowing day (O. Deir el-Medina 39 and P. Turin 1949 + 1946), as the 
accession took place in Thebes. For the accession dates of Ramesses 
IV, V and VI see Janssen, Varia, ch. VIII. Dates extend to 23/111/ 
Akhet year 7 (O. Deir el-Medina 207: KRI VI 149), and the Turin 
Indictment Papyrus (RAD 80,8-9) indicates that Ramesses IV must have 
died before the harvest of his year 7, i.e., before May (Gregorian), 
probably in IV Peret. The estimated age of the mummy (CG 61041) 
was at least 50 years. 

Ramesses V Beckerath calculated the accession at around 1/IV/ Peret 
[ZAS 122, 1995, 98; GM 157, 1997, 7-10); Gutgesell (Die Datierung der 
Ostraka und Papyri, Hildesheim 1983, 227-229), followed by Helck (Ostraka, 
411), prefers 29/111/ Peret, and Janssen (GM 155, 1996, 62) tries to 
set the date between 29/IV/ Peret and 7/1/ Shemu, but this remains 
uncertain. Recently, Beckerath put the accession before I Peret (GM 
188, 2002, 16-17), supported by Demaree's new interpretation of P. 
Turin 2044. 

O. Cairo CG 25247 (without royal name) documents 25/11 to 17/IV/ 
Shemu of year 4; a year 5 is not attested. The accession date of his 
successor indicates that he will have died between the middle of I Peret 
and the start of II Peret of his year 4. His mummy (CG 6 1 042) shows 
him to be the victim of a smallpox epidemic, who was probably just 
over 30 years of age (Harris & Weeks, X-raying the Pharaohs, London 
1973, 166-167). Krauss assigns him a reign of 3 years 10 months (OL£ 
90, 1995, 249). 

Ramesses VI His accession can be restricted to between 28/1 and 11/11 
Peret (Beckerath, GM 79, 1984, 7, based on O. IFAO 1425); argu- 
ments for 8/11/ Peret are given by Janssen (Varia, 131 — 138). The 


announcement followed on ?/II/ Peret (KRI VI 36 4). The highest date 
of the reign is 1 1 /II/ Peret year 8 ( Janssen, "Year 8 of Ramesses VI 
Attested", GM 29, 1978, 45—46), and a reign of 7 years 9 months can 
be deduced from Theban graffito 1860a (Krauss, OZ^ 90, 1995, 249). 
The earlier literature reflects confusion about the sequence (and thus 
names, numbers, dates and chronology) of the next two kings, but it 
is now clear that Ramesses VII (Itiamun) preceded Ramesses VIII 

Ramesses VII. (Itiamun) Suggestions for the accession date vary widely: 
from between 20/11/ Peret and 5/11/ Shemu (Janssen, JEA 52, 1966, 
92), to between 30/111 and 26/IV/ Peret (Beckerath, £i'S 122, 1995, 
99; but idem, Chronologie JVR, 85: on 30/111 or 1/IV/ Peret), between 
10/111 and 26/IV/ Peret (Janssen, GM 155, 1996, 61), and "shortly 
after" 14/IV/ Peret (Demaree, GM 137, 1993, 52). Utilizing P. Amiens 
and other documents, Beckerath recentiy proposed 30/111/ Peret (GM 
188, 2002, 17). Year 7 is documented in P. Turin 1907/1908 {KRI VI 
403-409) and O. Strasbourg H 84 with 16/11/ Shemu (Janssen, JEA 
52, 1966, 91 n. 2). While Eyre argues for a 25/IV/ Shemu year 8 
(P. Turin 1883 + 2095: "The reign-length of Ramesses VII", JEA 66, 
1980, 168-170), Valbelle suggests that the king in question may actu- 
ally be Ramesses IX [Owners, 39 with n. 8 and 204). 

Ramesses VIII The accession date has been fixed between 2/1/ Peret 
and 13/1/ Akhet (Amer, GM 49, 1981, 9-12) and, according to Beckerath 
(in: Deir el-Medina in the third millenium AD, Leiden 2000, 4-5), must lie 
before 13/1/ Akhet. Only year 1 is documented, but the calendar pro- 
vides a basis for arguing a year 2 if he reigned a full year, as his death 
must have occurred on 20/1/ Akhet because of the accession date of 
his successor. 

Ramesses IX His accession probably occured on 2 1 /I/ Akhet (Beckerath, 
GM 79, 1984, 7-8), but in any case on or after 18/1/ Akhet (Beckerath, 
ZAS 127, 2000, 112; Janssen, GM 191, 2002, 59-65). He died in his 
year 19 with III and IV Akhet documented (Botti, JEA 14, 1928, 48-51, 
without royal name), and Beckerath suggests more precisely the end of 
I Peret (ZAS 127, 2000, 112). 

Ramesses X The accession was between 25 and 27/1/ Peret (Beckerath, 
GM 79, 1984, 8-9, based on Turin Cat. 2075 + fragm. = KRI VI 653, 


and ZAS 127, 2000, 112, followed by Valbelle, Ouvriers, 43, n. 3, and 
Demaree, GM 137, 1993, 50). IV Akhet year 3 is documented in the 
Giornale, followed by 24/1/ Akhet year 1 (Helck, GM 70, 1984, 32). 
Krauss allows for a year 4 (GM 70, 1984, 42—43), but a year 8 pro- 
posed by Bierbrier (JEA 61, 1975, 251) is quite uncertain, and Parker's 
astronomical arguments for 9 full regnal years can be discarded (Krauss, 
Sothis, 151-154, and Jansen-Winkeln, £& 119, 1992, 32-33). 

Ramesses XI According to Helck (SAK 17, 1990, 211-212), his acces- 
sion date was 27 or 28/IV/ Shemu, but 20/111/ Shemu according to 
Gardiner (JEA 26, 1940, 23-29) and others (Cannuyer, Studies Lichtheim 
I, 98-105; Ohlhafer, GM 135, 1993, 59-72; Demaree, GM 137, 1993, 
50), but not between 26/111/ Shemu and the 4th epagomenal day 
(Beckerath, 0Z^85, 1990, 657 n. 1). His year 27 is documented (8?/IV/ 
Shemu: KRI VI 701,15). The Renaissance (whm-mswt-) era counted 10 
years and thus lasted into his 28th year (25/1/ Shemu year 10 in the 
letter BM 10326 = LRL no. 9). This means that a 29— year reign as 
given in the Book of Sothis is possible. Dates from the Renaissance 
era are listed by Thijs, GM 173, 1999, 190-191. Years of this era are 
to be correlated with the regnal years of Ramesses XI and do not 
accord with calendar years (Ohlhafer, GM 135, 1993, 59—72). Thijs' 
proposal that Ramesses IX— XI were contemporary (GM 167, 170, 173; 
SAK 31, 2003, 289-306) has been countered by Beckerath {gAS 127, 
2000, 114-116, and GM 181, 2001, 17). 


Karl Jansen- Winkeln 

At the beginning of Dyn. 2 1 Egypt was split in two, with two centres 
of power, each ruled individually. UE, whose northern frontier was 
located in the region of Herakleopolis, was governed by a military com- 
mander who, at the same time was HPA of Thebes. 1 In texts and 
depictions some of these UE regents (Herihor, Pinudjem I and 
Menkheperre) assume in varying degrees attributes which are reserved 
for a king. Kings reigned in LE, but at least two of them (Psusennes 
and Amenemope) occasionally bear the title of "HPA". Contemporaneous 
documents of which only a small number survived do not give any 
direct indication as to the reason for this partition of Egypt. 2 The only 
large group of finds are the graves of the kings in Tanis and the col- 
lective interments in the Theban necropolis (including replacements and 
re-interments of older mummies). Among these Theban funeral sites 
various dated objects can be found, but unfortunately most dates are 
anonymous and not ascribed to any explicit regent. Of this twofold 
line of regents, Manetho lists only the kings of LE, namely (1) Smendes, 
(2) Psusennes [I], (3) Nepherkheres, (4) Amenophthis, (5) Osochor, (6) 
Psinaches, (7) Psusennes [II]. Contemporary documents contain ample 
reference of the kings Psusennes (P>-sb>-h c j-m-nwt; only in LE), Amenemope 
{Jmn-m-Jpi) and Siamun (^-Jmri) (both in LE and UE). The first two 
kings can be straightforwardly identified as Manetho's Psusennes (I) and 
Amenophthis. A king named Smendes (JVs-b>-nb-ddi) is attested by only 
a few, undated inscriptions, but the history of Wenamun shows clearly 
that he was a contemporary of Herihor and thus the first king of Dyn. 
21. The identification of the remaining four kings, on the other hand, 
has caused some problems. 

The Nepherkheres of Manetho is not attested as the personal name 
of any king. Two bow caps from the grave goods of Psusennes I dis- 

1 The first two rulers also called themselves Viceroys of Nubia; the first three had 
the title Vizier. 

2 For an attempted explanation, see K. Jansen-Winkeln, Orientalia 70 (2001), 153-182. 

DYNASTY 21 219 

play the throne-name and the personal name of Psusennes opposite the 
throne-name Nfr-k'-R? (hq? Wist) and the personal name Mrjj-Jmn Jmn- 
m-njswt? Obviously the throne-name Nfr-kS-R' has been handed down 
as Nepherkheres by Manetho. The proper name Amenemnisut (Jmn- 
m-njswi) is attested only a second time on the relief Berlin 23673 from 
the reign of Shoshenq V, on which a long line of ancestors of the 
owner is named, sometimes together with the reigning king. On this 
relief, Amenemnisut is the predecessor of Psusennes I, whereas Manetho 
names him as the successor. His true position has not yet been identified. 
The Berlin genealogy was compiled only about 250—300 years after 
the reign of Amenemnisut and should, therefore, be given greater con- 
sideration than Manetho's frequently garbled tradition. But the fact that 
Psusennes and Amenemnisut appear together on one funeral object, 
strengthens the idea that Amenemnisut was the successor of Psusennes 
and that he donated the object. 4 Nevertheless, Amenemnisut (Nepher- 
kheres) was without question an ephemeral king. 

The Osochor of Manetho is attested contemporarily only by one 
inscription from Karnak, which registers the inauguration of a priest 
in year 2 of a king with the throne-name 'l-hpr-K Stp.n-R c (the personal 
name is missing in a lacuna).' 1 E. Young has demonstrated 6 that this 
king cannot be Psusennes I, as believed in the past, because he always 
bears the epithet Stp.n-Jmn. Furthermore, a few lines further down, the 
text refers to the inauguration of the priest's son in the year 17 of 
Siamun. If 'l-hpr-'l Stp.n~R c really were identical with Psusennes I, then 
the inaugurations of father and son had to have been almost three gen- 
erations apart. Therefore this otherwise unknown throne-name from 
Dyn. 2 1 may well be that of Manetho's Osochor. In this case the sec- 
ond inauguration would have taken place only 21 years, or about one 
generation, later, if Manetho's 6 years for Osochor be accepted. 

Actually, the personal name of the king is mentioned once, but not 
contemporarily. 7 An inscription (no longer traceable) from the roof of 

3 Montet, Tanis II, 105; 108, Fig. 44; pi. 72 (No. 413/414). 

4 See also Kitchen, TIP, 70-71. An alternative to this could be that the bow was 
made during a co-regency of the two kings, cf. TIP, 70—7 1 and Beckerath, Chronologic, 
101. However the reign of Nepherkheres only lasted for a few years, and a co-regency 
is more likely at the end of a long reign. For the Berlin genealogy see Bochardt, Mittel, 
96-112; Bl. 2/2a. 

5 No. 3B of the "Annals of the Priests", see G. Legrain, RT 22 (1900), 53; Kruchten, 
Annates, pi. 2; 17. 

6 JARCE 2 (1963), 100-101. 

7 Concerning the following see J. Yoyotte, BSFE 11-1% (1976/77), 39-54; cf. also 
Kitchen, TIP, § 437. 


the temple of Khonsu from year 9 of Takelot IIP mentions, among 
the author's ancestors, a king Osorkon and his mother Mhjt-m-wsht. 
This Osorkon cannot be identical with one of the kings named Osorkon 
from Dyns. 22—23, because their mothers had different names. 9 A king's 
mother called Mhjt-m-wsht is known from Dyn. 2 1 ; on the stela of P>- 
sn-Hr from the Serapeum the grandmother of Shoshenq I is named 
likewise. 10 The two texts complement each other optimally and indi- 
cate the existence of a king Osorkon in Dyn. 2 1 , the uncle of the later 
Shoshenq I, who can be identified as Manetho's Osochor. They also 
match in time: Psusennes II, father-in-law of Osorkon I, is assumed to 
have been a contemporary of Shoshenq I. Because "Osochor" 11 was 
the older brother of Shoshenq I's father, he might well have been the 
second predecessor of Psusennes II. 12 The identification of Manetho's 
Osochor by Young and Yoyotte has gained general acceptance. 

Manetho's last king but one, Psin(n)aches, cannot be found in any 
Egyptian sources whatever. The only name that could be considered 
(with some modifications), would be Pj-sb>-hj-m-nwt, r '' but that name 
has already, and rightly so, been identified as Psusennes. On the other 
hand, contemporary documents reveal a King Siamun {/^-Jmri) bear- 
ing the throne-name Mrj-hpr-R c as the last but one king of Dyn. 21, 
who does not appear in Manetho's history. It is tempting, therefore, 
to identify Manetho's Psinaches with Siamun, 14 even though the lengths 
of their reigns do not match: Manetho's Psinaches is supposed to have 
reigned for nine years, Siamun, by contrast, for at least 17 years. A 
solution would be to amend the number 9 to <1>9. This identification 
and emendation have become traditional, as the most obvious. Anyway, 
we should always bear in mind that this identification originates only 

8 ID, III, 258c; G. Daressy, RT 18 (1896), 51-52. 

9 Nor can Osorkon III and IV be considered, for chronological reasons. 

10 See CSSM, 30-31; Kitchen, TIP, § 85. 

11 This form of the Egyptian-Libyan word Ws(j)rkn ("Osorkon") is attested elsewhere 
in Manetho. 

12 F. Payraudeau, "Remarques sur l'identite du premier et du dernier Osorkon", 
GM 178 (2000), 75-80, is of the opinion that two objects of a king c ,'-hpr-R' stp.n-Jmn 
Mrjj-Jmn Wsjrkn, which until now have been ascribed to Osorkon IV (whose throne- 
name is unknown), originally belonged to Osochor. If this is correct, Osochor would 
have taken turns using the epithets stp.n-R' and stp.n-Jmn in his throne-name. 

13 Cf. M. Romer, GM 114 (1990), 94. 

14 Cf, most recendy, J.v. Beckerath, GM 130 (1992), 17-19 and (concerning a pos- 
sible explanation for the varying information in Manetho's work) GM 131 (1992), 11. 

DYNASTY 21 221 

from the fact that we can neither find a king from the end of Dyn. 
21 who is named in contemporaneous documents in Manetho's work, 
nor can we find Manetho's last but one king Psinaches on Egyptian 
monuments. The remaining two criteria for the identification have not 
been met: neither name nor length of reign being the same. 

The identification of Manetho's second king called Psusennes with a 
(Hr-) P)-sb)-h c j-m-nwt is, on the one hand, unequivocal and undisputed. 
On the other hand, however, there is the question as to whether the 
last king of Dyn. 21 is identical with the last HP of Thebes of that 
dynasty who has the same name. 1 ' 1 Actually, the evidence weighs heav- 
ily in favour of his being one and the same man, who was first HP 
and then successor to King Siamun in Tanis, without giving up his 
Theban office. 

The only reference for the HP Psusennes can be found on shrouds 
and mummy-braces (etc.) from the priests' mummies in the so-called 
second Cachette (Bab el-Gusus). 16 From 10 references, 8 name him 
HP, whereas on the other 2 1 ' his name appears in a cartouche. No 
other titles are mentioned, which for H. Kees meant that he — in con- 
trast to his predecessors — no longer possessed military power. 18 But this 
conclusion was perhaps overly hasty, because the HP Menkheperre, 
who held the highest offices, 19 is referred to on mummy wrappings from 
the second Cachette as only a HP, 20 his name otherwise appearing in 
a cartouche; 21 his military titles are not mentioned at all, and in the 
filiations of his descendants his name is often cited without any titles. 22 
It can be established that the HP Psusennes' name is sometimes written 
in a cartouche like the names of Herihor, Pinudjem (I) and Menkheperre, 
whilst his father and predecessor Pinudjem II never used any royal 

15 In Kitchen, TIP they are distinguished from one another as Psusennes II (= the 
king) and III (= the HP). 

16 Burials A. 17; 43; 48; 58; 65; 66; 125; 132; 133; 148, see G. Daressy, ASAE 8 
(1907), 23-37. 

17 A.58 and 66, see Daressy (n. 16). 

18 H. Kees, Die Hohenpriester des Amun von Karnak von Herihor bis zum Ende der Athiopen- 
zeit (Leiden: PA 4, 1964), 79: "In contrast to all of his predecessors in Thebes he did 
not style himself supreme commander of the UE army." 

19 Cf. M. Romer, Gottes- und Priesterherrschqft am Ende des Neuen Reiches (Wiesbaden: 
AUAT 21, 1994), 66-73. 

20 Burials A.2; 13; 96; 105; 109; 113, see Daresssy (n. 16), 22-31. 

21 A.ll and 64, Daressy (n. 16), 22; 27. 

22 A. 12; 26; 32; 38; 81 (Daressy [n. 16], 22-28) and elsewhere. 


attributes. A graffito from the Temple of Abydos 23 reveals the com- 
plete titles of a king Tjt-hpr-K 1 Stp.n-R c P>-sb>-hj-(m-)nwt Mrjj-Jmn, who 
is simultaneously HPA and supreme military commander. Actually, the 
elements of the titles of his kingship, his duties as HP and his military 
titles blend into each other in a peculiar manner, not to be found else- 
where. He is called 

njswt-bjt nb t>wj Tjt-hpr-R c Stp.n-R c <mrjj>(?) Jmn-R c njswt ntrw 2 ' 1 

hm-ntr tpj n Jmn-R c njswt ntrw 

z> R c nb hjw 

hlwtj P?-sb?-hj-(m-)nwt Mrjj-Jmn ntj (r-)hlt nl ms c [w n Knit drwj 

hm-ntr tpj n Jmn-R c njswt ntrw 

jrj hpw nfrw n Kmt 

hHwtj pr-> P>-sbj-hj-(m-)nwt Mrjj-Jmn. 

The military title, hlwtj pr J > P>-sb>-hj-(m-)nwt ntj (r-)h>t n> ms'[w n Kmt 
drw], is very informative. It reveals distinctly that this is the HP Psusennes, 
the successor of Pinudjem II, and not a king who has adopted the 
additional title of HP (as Psusennes I and Amenemope did). The rea- 
son is that this title is only to be found in connection with Theban 
HP and military commanders, 23 but never in connection with a Tanite 
king. The throne-name of Psusennes in this graffito also appears with 
slight variation (Tjt-hprw-R c ) on a vessel fragment from Abydos. 2b A king 
bearing almost the same name, Tjt-hpr-R' Stp.n-R' Mrjj-Jmn Hr-P>-sb>- 
hj-m-nwt, can be found outside Abydos on two Theban statues: (1) 
Cairo CG 42192, on which he is named as an ancestor of his grand- 
son M c -hpr-R c Stp.n-R c Mrjj-Jmn Ssnq (Shoshenq II); 27 (2) the Nile-statue 
London BM 8 of that particular grandson which also mentions the 

23 M. A. Murray, The Osireion at Abydos (London, 1989), 36; pi. XXI; G. Daressy, 
RT 21 (1899), 9-10. 

21 Concerning this epithet, cf. M.-A. Bonheme, Les norm royaux dans t'Egypte de la 
Troisieme Periode Intermediaire (Cairo: BdE 98, 1987), 61. 

25 See GM 99 (1987), 19. No. 8 is to be crossed out of this list, seeJEA 81 (1995), 
130; instead, the HP Jwlt is attested a second time on an altar-stand in Moscow, see 
S. Hodjash & O. Berlev, The Egyptian Reliefs and Stelae in the Pushkin Museum of Fine Arts, 
Moscow (Leningrad, 1982), 157/161 (No. 105). 

26 E. Amelineau, Les nouvelles fouilles dAbydos 1897-1898 (Paris, 1904), 146 (24). 

27 Shoshenq II donated the statue, and not Schoschenq I, as often reported; cf. J.v. 
Beckerath, Orientalia 63 (1994), 84-87 and K. Jansen-Winkeln, JEA 81 (1995), 145-148, 
who both render the text. 

DYNASTY 21 223 

daughter of Psusennes and mother of Shoshenq II, A4>'t-k>-R c .' 2? ' The 
additional Hrf-Pf-sk-h'j-m-nwt) 29 is not a distinctive feature, but appears 
with reference to one and the same person 50 as demonstrated by the 
Decree for Maatkare. 31 Here the very same Psusennes and father of 
Maatkare is solely called Mrjj-Jmn P?-sb?-h c j-m-nwt. The obvious con- 
clusion is that all these cartouches refer to the same person. 32 The 
graffito from Abydos also demonstrates that he was king and at the 
same time HP in Thebes; he had clearly not resigned this office. 33 He 
was probably buried in Tanis (and later re-buried in the Antechamber 
of the tomb of Psusennes I). 34 A limestone-fragment with his name has 
been found near Tell el-Daba. 35 

Dodson drew the conclusion that Psusennes did not have a reign of 
his own at all, but was only an UE ephemeral King next to Sho- 
shenq I, 36 from the fact that many of his attestations are posthumous 
and that he is often mentioned together with Shoshenq I. 37 This is not 
at all convincing: on CG 42194 and BM 8 he is only mentioned in 
his grandson's genealogy, and together with Shoshenq I he only appears 
in the tomb TT A. 18. 38 By contrast the latter inscription provides 

28 C. R. Lepsius, Auswahl der wichtigsten Urkunden des Aegypischen Alterthums (Leipzig, 
1842), pi. XV. 

29 Another reference is an inscription on a bead of unknown provenance, see GLR 
III, 300 (IV). The ivory stick-handle, which in Gauthier's opinion also belonged to 
Psusennes II (GLR III, 302 [IV]), more probably belonged to Psusennes I. 

30 In contrast to Bonheme (n. 24), 60, who inexplicably would like to recognize up 
to four different persons in Tjt-hpr-R' (Hr-)Pj-sb'-h'j-m-nwt; cf. also J. Yoyotte, BSFFT 
1 (1988), 46(1). 

31 J. Winand, Cahiers de Karnak XI (2003), 672ff; 707 (Fig. 4), 1.3, 5. 

32 Beckerath's distinction (GM 130 [1992], 18) between a Tanite king (Hor-) Psusennes 
with the throne-name Tjt-hprw-R' Stp.n-R' and a HP who, in the role of a (mock-) king 
(on the graffito in Abydos) bore the throne-name Tjt-hprw-R' Stp.n-Jmn is not correct. 
The epithet is Stp.n-R' in this graffito, too. Furthermore he is also called Tjt-hprw-R' 
Stp.n-R' on a vessel fragment from Abydos (n. 26, above) which cannot be connected 
to any other Psusennes than the one from the graffito. 

33 It is inexplicable why Beckerath (GM 130, 1992, 18) writes that if the HP Psusennes 
had inherited the crown from Siamun he would have had to appoint a new HP. A 
HP's and a king's office do not exclude each other in dynasty XXI. 

34 Cf. Yoyotte (n. 30), 41-53; idem, Tanis, L'or des pharaons. Exhibition-catalogue 
(Paris, 1987), 136-137. 

3: ' M. Bietak, Avaris and Piramesse: Archaeological Exploration in the Eastern Nile Delta 
(Oxford, 1981), 271; but cf. LA V, 131. 

36 RdE 38 (1987), 49-54; BES 14 (2000), 9-12. 

37 The statues Cairo CG 42192; CG 42194 (name destroyed), London BM 8 and 
the Decree for Maatkare. 

38 A. Dodson, JEA 79 (1993, 267-268; pi. 28. On Cairo CG 42192, on the other 
hand, Psusennes does not appear together with Shoshenq I, but with another king 
called Shoshenq, see above, footnote 27. 


weighty evidence that Shoshenq I was Psusennes's successor: someone 
is promoted by Psusennes (shntj.f), and is promoted once again during 
the reign of Shoshenq (whm hntj.f). There is no reference that the two 
kings reigned in parallel. Considering the fact that Psusennes II was 
buried in Tanis and appears in Manetho's list of kings, we cannot say 
that he was only an UE ephemeral king/ 9 

There is thus evidence of the following LE kings in Dyn. 21: (1) 
Smendes, (2/3) Psusennes/Amenemnisut (Nepherkheres), (4) Amenemope, 
(5) Osorkon ("the Elder", Osochor), (6) Siamun ("Psinaches" in Manetho's 
work) and (7) Psusennes II (at the same time HP in Thebes). We do 
not know much about the familial relationships of these kings. There 
is nothing to be said about the origin of Amenemnisut, Amenemope 
and Siamun. Smendes I's wife, Tentamun, 40 is mother of Henuttawy, 
wife of Pinudjem I and mother of Psusennes I. 41 Consequently Smendes 
I would be father-in-law of Pinudjem I and grandfather of his succes- 
sor(?) Psusennes I. Osochor, being son of Nimlot I and of Mhjt-m-wsht, 
is uncle of Shoshenq I. Psusennes II is son of HP Pinudjem II and 
father-in-law of Osorkon I; the latter already concerns the relationship 
to the kings of Dyn. 22. 

Nine UE rulers are known as belonging to Dyn. 2 1 . Eight of them 
are part of a lineage of fathers and sons: 


Pinudjem I 

Masaharta Djedkhonsiuefankh Menkheperre 

Smendes II Pinudjem II 

Psusennes (II =) III 

39 Cf. also Beckerath, GM 130 (1992), 17f; Kitchen, 77P 3 1995, XIX-XXI. 

40 Herself being the daughter of a man without any important titles, called Nebseni. 

41 Under the probable circumstance that the King's Mother who is mentioned on 
some funerary objects, Henuttawi is identical with Pinudjem's wife. Kitchen's postu- 

DYNASTY 21 225 

Only Herihor does not belong to this lineage; 42 his position as a pre- 
decessor or successor of Payankh is the only one disputed (see below). 
The order of the others is clear, even though some overlap. There is 
evidence of the HP (and king) Pinudjem I until a year 15, and of his 
son Masaharta in the years 16 and 18 following. His son Menkheperre 
who is clearly younger takes up the duties of the HP in a year 25 and 
from then holds office for almost five decades. A third son of Pinudjem 
I, called Djedkhonsiuefankh, is recorded only once as a HP on a coffin 
which is at present missing; 43 he most probably held office for a very 
short time between Masaharta and Menkheperre. 44 However, Pinudjem 
survived his son's term of office and died in that of Menkheperre (see 
below). Evidence of Smendes II is, admittedly, somewhat better than 
that of Djedkhonsiuefankh, but his term of office can only have been 
very short, either as Menkheperre's successor or as his "co-regent" (see 
below). His brother Pinudjem II came next in office, followed by his 
own son Psusennes, who is probably identical with King Psusennes II 
(see above). Consequently we have the following order: (1/2) Payankh and 
Herihor (see below), (3) Pinudjem I, (4) Masaharta, (5) Djedkhonsiue- 
fankh, (6) Menkheperre, (7) Smendes II, (8) Pinudjem II, (9) Psusennes III. 

Some of the UE regents are related by blood or marriage to those 
of LE: Smendes I seems to be the father-in-law of Pinudjem I, Pinudjem 
himself is Psusennes I's father (see above). HP Psusennes himself becomes 
king in Tanis. 

Concerning the succession of the first two HP, Herihor was on 
account of a copying error believed for a long time to be father and 
predecessor of Payankh. Since this error has been corrected, 45 

late of a second (older) Henuttawi "Q_" as a hypothetical second wife of Smendes and 
Psusennes' mother is only rooted in his wish for a genealogical bridge to the Ramessides 
for Psusennes on account of his occasionally being called "Ramses-Psusennes". This 
has only confused matters unnecessarily. 

42 Depending on whether he was predecessor or successor of Payankh, he might 
have been his father-in-law (Kitchen, TIP 1 § 438) or son-in-law (K. Jansen-Winkeln, 
ZAS 119 [1992], 25) or he might have married his wife after Payankh died (J. Taylor, 
in: Eyre, Proceedings, 1143—1155). 

43 Kitchen, TIP § 392. 

44 We cannot totally exclude the possibility that he was a predecessor of Masaharta's 
who was in office only for a short period. According to A. Niwinski (BES 6, 1984, 
83-6) he was a son of Pinudjem II; Torr's filiation data would in consequence not 
name his father but his great-grandfather (!) Pinudjem I. 

4:> Cf E. F. Wente, (Fs Korostovtsev) Drevny Vostok (Moscow, 1975), 36-38; The Temple 
o/Khonsu. I. OIP 100 (1979), p. 13(d); pi. 26, 1. 4. 


the succession has had to be explained by other means. The term of 
office of both HP or at least part of it can be said to have taken place 
in the later years of the reign of Ramses XL Records mention Herihor's 
years 5 and 6 (without any explicit relation), and Payankh's year 7 of 
the whm-mswt-era and a year 10. At first sight it would be logical if 
Herihor had held office in the first half of the whm-mswt-era and Payankh 
in the second. Even so, a series of arguments favour a reverse order: 46 
(1) The form of the titles: We can recognise Payankh's origin from the 
rank of officers much more clearly than that of Herihor. He is mostly 
referred to simply as "The General", his military titles being much 
more prominent and detailed than those of Herihor. His titles are in 
general similar to those of Pinhasi, who was in charge of UE from the 
beginning of the whm-mswt-era.. The titles of Herihor on the other hand 
are more related to those of the later HP. Furthermore, Payankh's titles 
almost always refer to the king (. . . n pr J >), as was usual in the Ramesside 
period, whereas those of Herihor no longer do so. (2) Payankh never 
assumes any royal titles or attributes, whereas Herihor and the later 
HP do. (3) Herihor and Pinudjem I are both recorded as builders in 
Thebes, and Pinudjem directly succeeds Herihor with regard to the 
decoration of the temple of Khonsu. Payankh on the other hand is not 
recorded as a builder. A similar situation is to be found regarding the 
(re-)burials in the Theban necropolis. On shrouds, bandages etc. of 
these mummies, every single HP of Dyn. 2 1 is recorded, except Payankh. 
Thus these burials must have taken place after his term of office. (4) 
The genealogical information corresponds more to a Payankh-Herihor 
succession. The order of these HP is still being discussed, 4 ' but in my 
opinion the order Payankh-Herihor is the more probable solution. At 
any rate, this problem has a direct influence on the chronology of the 
whole dynasty. 

Most of the dates preserved from Dyn. 21 are from Thebes, and 
most do not refer to a specific ruler. Breasted presupposed that all 

46 %AS 119 ( 199 2), 22-25. 

47 The following authors do not agree with the thesis published in J(Ag \ \ 9 naming 
Herihor as Payankh's predecessor: A. Niwinski, BIFAO 95 (1995), 346-47; J.v. Beckerath, 
in: D. Kessler & R. Schulz, eds., Gedenkschrift fur Winfried Barta (Frankfurt: MAU 4, 
1995), 49-53; A. Gnirs, Militdr und Gesellschaft (Heidelberg: SAGA 17, 1996), 199-201; 
Kitchen, TIP 3 1995, XIV -XIX (A-N). For a response see K. Jansen-Winkeln, GM 
157 (1997), 49-74. In favour of the succession Payankh — Herihor are A. Egberts (GM 
160 [1997], 23-25; £4$ 125 [1998], 93-108) and J. Taylor (see above, footnote 42). 

DYNASTY 21 227 

those dates were related to the LE kings 4 " and in recent times this 
opinion has found general acceptance. The opinion is supported by 
some explicit dates which almost always mention the name of a Tanite 
King: There is a date which is explicitly related to Amenemope, 49 
another one is related to Osochor,'' and six are related to Siamun.' 1 
Furthermore Amenemope and Siamun are quite well documented in 
Thebes. On the other hand there is only one date which is explicitly 
related to a HP. 52 Under that condition, the following years would be 
recorded: 53 

Whm-mswt-era: 4; 5; 6; 7; 10 

Smendes I: 1; 4(?); 6; 9; 10; 11; 12; 13; 15; 16; 18; 19; 20; 21; 25 

Amenemnisut: — 

Psusennes I: 6; 7; 8; 19; 27; 54 30; 40; 48; 49 

Amenemope: 1; 3; 5; 10(?) 55 

Osochor: 2 

Siamun: 1; 2; 3; 5; 6; 7; 8; 9; 1_0; 12; 14; H3; 17 

Psusennes II: 5; 13(?) 56 

The currently recorded dates can be made compatible in this order 
with Manetho's lengths of reign as preserved by Africanus. He gives 
Smendes 26 years, Nepherkheres (Amenemnisut) 4, Psusennes 46 (41 
according to Eusebius), Amenemope 9, Osochor 6, Psinaches (that 
means Siamun [?]) 9 and Psusennes (II) 14 (35 according to Eusebius). 
At the end of Psusennes I's reign there was supposedly a co-regency 
with Amenemope. On condition that this dating system was used, the 
famous linen-bandage with the inscription "King Amenemope; year 49" 
can be restored beyond doubt to "[year X under] King Amenemope; 
year 49 [under King Psusennes; linen made by HP NN . . .]", 

48 BAR IV, 

§ 604 


49 Kitchen, 





50 Kitchen, 





51 Kitchen, 

TIP § 




73; 74; 77; 82 (from LE); 83; 84. 

52 Kitchen, TIP § 387, no. 46. This latter record must be interpreted differently if 
we adhere to a continuing dating by LE kings, cf. e.g., E. Young, JARCE 2 (1963), 
102-103, n. 21; Kitchen, TIP § 377. 

53 With reference to the evidence listed in Kitchen, TIP § 379-381. Only the under- 
lined dates are connected explicitiy with the king. 

54 A. Dodson & J. J. Janssen, JEA 75 (1989), 128, 134. 

55 This date may also refer to Siamun, cf. Kitchen, TIP § 388, no. 55. 

56 Possibly referring to Shoshenq I, cf. Kitchen, TIP § 391, no. 86; 87. 


the juxtaposition of the years being evidence of a co-regency. 37 As a 
result, Amenemnisut must have been the predecessor of Psusennes, and 
the only contemporary record of this king would indicate a co-regency 
Amenemnisut — Psusennes 58 at the beginning of Psusennes I's reign. 
Various suppositions have been made concerning the length of these 
two (hypothetical) co-regencies, 59 almost all of them deriving from 
Manetho's information: Only 46 of the 49 recorded years were to be 
taken into consideration. 

On the other hand, in the case of Psinaches/Siamun, Manetho has 
to be emended. Siamun's attested 17 years mandates the emendation 
9 > 19 (0 > 10). Altogether Dyn. 21 would have lasted 124 years which 
is the result of adding the lengths of reign according to Africanus and 
this emendation. The difference between these 124 years and Manetho's 
sum of 130 years (indicated in all versions, regardless of the actual, 
correct total) might be explained by suggesting that Manetho calcu- 
lated those years in which there was a co-regency for both rulers. 50 
According to this hypothesis, the lengths of reign for the UE rulers 
would be as follows: 

Herihor until year 6 (or 7) of the whm-mswt-era; 

Payankh from year 6 (or 7) until year 1 of Smendes I at most; 

Pinudjem at the earliest from year 10 of the whm-mswt-era. onward, 

until year 15 (year 16 at most) of Smendes I in his position as HP, 

after that at least until year 8 of Psusennes I as king; 

Masaharta from year 16 (15 at the earliest) until year 25 of Smendes 

I as a HP at the latest; 

Djedkhonsiufankh only for a very brief period between Masaharta 

and Menkheperre; 

Menkheperre from year 25 of Smendes I until (at least) year 48 of 

Psusennes I. 

Smendes II for a brief period between Menkheperre and Pinudjem II; 

Pinudjem II from year 1 of Amenemope or shortly thereafter; 61 

Psusennes "III" from year 10 of Siamun on. 

If, however — which seems probable — Payankh is not the successor but 
the predecessor of Herihor, this system cannot easily be maintained. 

Kitchen, TIP § 29. 

See above, footnote 4 and Kitchen, TIP § 56. 

Cf. for example Kitchen, TIP I 29, 465 (table I) or Beckerath, Chronologic, 101-102. 

Cf. Beckerath, Chronologie, 101-102; idem (n. 47), 54-55. 

Kitchen, TIP § 388 (51). 

DYNASTY 21 229 

The highest recorded date for Herihor is a year 6, 15/111/ Peret. 62 This 
date could only refer to Smendes if Herihor followed Payankh and if 
the dates refer exclusively to the LE kings. But Pinudjem was already 
recorded in year 6, 7 /III/ Peret (of Smendes after this system). The above 
given dating-system could only be retained if Herihor's date was to be 
read 7 /III/Akhet hi instead of 7 /III/ Peret (or emended accordingly), but 
that would be an unhappy solution. 

Even so, there is some information for Theban dates of the UE 
kings. There is a record of a year 48 of HP Menkheperre; 64 moreover, 
a closer look reveals a complementary distribution of the records con- 
cerning the rulers of that time. In the first half of Dyn. 21, HP Herihor, 
Pinudjem I 6n and Menkheperre have royal attributes and titles to differing 
extents. On the other hand, the LE kings of that time are virtually not 
recorded at all in UE: there is a graffito mentioning Smendes 66 and a 
rock-stela, 67 and nothing for Amenemnisut and Psusennes I, even though 
the latter reigned for a long time. Subsequently, however, Amenemope 
and Siamun are well documented in Thebes, and Osochor at least 
once, whereas HP Pinudjem II (who held office parallel to them), does 
not adopt any royal attributes or titles. It is, therefore, likely that the 
HP who called themselves kings counted their own years of reign 
whereas during the second half of the dynasty the dates refer to the 
LE kings. This would mean that the beginning of Amenemope's reign 
might have implied a change in the dating-system and concurrently a 
change in the political structures. 68 

A possible, but very hypothetical explanation would be that a new 
family or a new branch of the same family gained power in Tanis and 

62 Kitchen, TIP § 379, no. 3 

63 Cf. ZAS 119, 26; Beckerath (n. 47), 51. 

64 Kitchen, TIP § 387, no. 46. 

65 This does not, of course, apply to Masaharta and Djedkhonsiuefankh since their 
period is equal to that of Pinudjem I. 

66 A. Varille, Kamak(-Nord) I (Cairo 1943), 36, Fig. 26, pi. 98 (71); L. A. Christophe. 
Karnak-Nord III (Cairo 1951), 77. 

'" G. Daressy, RT 10 (1888), 135f. Already in Daressy's time part of the text was 
gone; in the meantime everything has been destroyed. The genre of the text (Konigsnovelle) 
normally requires a date, but the structure of the text does not require a date in that 
part which was already missing in Daressy's time. 

68 P. Brooklyn 16.205 might contain some information with regard to a critical sit- 
uation in UE, referring to a year 49 of Dyn. 21 as a "bad time" (Ib'w bjn); concern- 
ing the dating of the papyrus to Dyn. 21, see J.v. Beckerath, GM 140 (1994), 15-17; 
Kitchen, TIP\ XXVI (Y). 


then successfully laid claim to supremacy over the whole of Egypt. We 
know that Smendes and Psusennes I were closely related to the UE 
family of HP (see above). No family relationships whatsoever are known 
for Amenemope and Siamun, but Osochor, who held office between 
them, was a son of the Libyan great chief of the Meshwesh, Shoshenq 
A, and the uncle of the later Shoshenq I. We do not know if this fam- 
ily was in any way related to the descendants of Payankh, although it 
is possible that Amenemope, Osochor and Siamun all belonged to this 
family, or to a branch of it. It is also striking that HP Pinudjem II, 
son of Menkheperre is not only called his son [z> Mn-hpr-R c ), but also, 
sometimes even on the same object, the son (= descendant) of King 
Psusennes (I). 69 Thus it seems to have been important to stress his being 
part of this half of the royal family. A change of royal family with 
Amenemope could explain a change within the dating-system. 

Assuming that the UE regents Herihor, Pinudjem I and Menkheperre 
counted their own regnal years, we can draw some conclusions. For 
the period of the LE kings Smendes I, Amenemnisut 70 and Psusennes 
I just one single date would have been recorded in Egyptian sources, 
and even that from later times: Year 19 of a king Psusennes is men- 
tioned in retrospect on a stela from the Dakhla Oasis dated to year 
five of Shoshenq (I). 71 Under these conditions only the regnal years as 
given by Manetho could be used as evidence for the dates of these 
kings — which is precisely what scholars have done. 

Uncertainty prevents us from precisely calculating regnal years for 
the first three UE rulers. Herihor reigned for at least 5 whole years 
(year 6 is recorded), possibly slightly longer (up to 8 years). In year 25 
of Pinudjem, his son Menkheperre was installed as HP, 72 and soon after 
that a new count of years begins.' 3 As a result we have to calculate at 

69 Daressy (n. 16), 23 (no. 24); 27 (no. 61); 28 (no. 81, no. 82); 31 (no. 113); 32 
(no. 119, 120); 36 (no. 139). 

70 If the linen-band with the regnal year 49 (cf. above) is not to be associated with 
Psusennes but rather with Menkheperre, there is no need to propose a co-regency for 
Psusennes and Amenemope. As a result, the question of whether Amenemnisut was 
predecessor or successor of Psusennes reappears (see above, footnote 4. 

71 A. H. Gardiner, JEA 19 (1933), 32; pi. VI, 1.11. Concerning the dating see 
H. Jacquet-Gordon, in: Hommages a la memoire de Serge Sauneron I (Cairo: BdE 81/1, 
1979), 180-182; O. Kaper, BACE 12 (2001), 77, n. 6; R. Krauss, DE 62 (2005), 43-48. 

72 "Banishment Stela," 1. 1-8, see J.v. Beckerath, RdE 20 (1968), 10. 

73 In line 7/8 of the Banishment Stela a lower date follows (RdE 20, 10-11; 33). 
The two events described in the text should not lie too far apart from each other. 

DYNASTY 21 231 

least 24 years for Pinudjem, at most 25 years. The highest date recorded 
for Menkheperre is the year 49 and in that year (his last?) Amenemope 
may have already reigned in Tanis. 74 

At first sight this seems to be contradicted by the fact that Smendes 
II, son of Menkheperre would have had to be HP at the latest when 
Psusennes I died, 75 because he donated goods for the burial. For this 
reason he cannot have been Menkheperre's successor if the reign of 
Menkheperre overlaps with that of Amenemope and even less so if Ame- 
nemnisut was Psusennes's successor. Niwinski presumed that Smendes 
II was only HPA in Tanis at that time, later becoming Menkheperre's 
successor for a short time. 76 This is possible, but in my opinion it is 
more probable that Smendes — like Masaharta previously — held office 
parallel to his father at the end of his father's reign, while the count- 
ing of regnal years continued to follow Menkheperre's reign. However 
that may be, 48 years is the most likely calculation for Menkheperre. 77 

Consequently, the first three UE rulers could be reckoned to have 
held office for at least roughly as long as the LE kings, namely 77 
years (5 + 24 + 48), possibly 1 or 2 years less, if the overlap between 
Menkheperre and Amenemope is greater. A slightly longer period seems 
to be more probable, including some leeway for Herihor, altogether 
perhaps 80 years, hardly significantly longer. In other words, the dates 
we have from Manetho's tradition, 124 years (the sum of the lengths 
of reign according to Africanus with emendation 9 to 19 for Psina- 
ches) and 130 years (sum total in all versions), set the limits of what 
is possible. Most likely is a total of about 126-8 years. If there is a 
difference in the lengths of the reigns of the HP Herihor, Pinudjem I 

The lower date is probably the first year of Menkheperre, who grants an amnesty on 
New Year, which is an appropriate act at the beginning of a reign. Thus the event 
would mark the transition from Pinudjem to Menkheperre, not only the inauguration 
of a new HP under the royal authority of Pinudjem which would have been implied 
by the other dating-system. 

74 The linen-band with the inscription "[Year X under] King Amenemope; Year 
49 [under NN]" (see above) under this circumstance would contain a common date 
of Amenemope and the HP Menkheperre. 

75 Cf. Kitchen, TIP § 25. 

76 A. Niwinski, JARCE 16 (1979), 59-60; idem, 21st dynasty Coffins from Thebes (Mainz: 
Theben 5, 1988), 50-51 (§ 43). 

77 Less only if the overlap with the era of Amenemope lasted longer. But if the 
"bad time" for year 49 (footnote 68) is connected with the change of regency to 
Amenemope, year 49 of Menkheperre could be the same as year 1 of Amenemope. 


and Menkheperre and the parallel reigning LE kings (from the reign 
of Amenemope onwards there is no difference in the two dating-sys- 
tems anyway) it would only amount to a few years. And we do not 
know if Smendes and Herihor started their reign at the same time or 
whether the Manethonian numbers are all correct. 78 

At the beginning of Dyn. 22 there is a certain fixed point which 
links Dyn. 21 to absolute chronology, i.e. Shoshenq I's campaign in 
Palestine. According to the OT,' 9 the Egyptian King Shishak besieged 
Jerusalem in year 5 of Rehabeam, king of Judah. On the Egyptian 
side, the campaign is attested by a victory scene in Karnak. Year 5 of 
Rehabeam can be pinned down to about 926/925 BC with the aid of 
the known lengths of reign of the kings of Israel and Judah and their 
synchronisms — although there are some inconsistencies — as well as by 
means of two synchronisms with the Assyrian chronology. 80 

From Egyptian sources we do not know when Sheshonq's campaign 
took place. Construction work on the pylon and the court, on whose 
exterior walls the scene of triumph is depicted, began in his year 21 
(possibly his last year but one), as recorded on a rock-stela. 81 The major- 
ity opinion is that the construction work and the campaign were con- 
nected to each other and that the campaign did not take place very 
long before construction work started, in year 20 at the earliest. An 
essential point for the temporal connection between the campaign and 
the construction work could be that of the whole decoration which was 
planned in Karnak only this triumphal scene has been completed. So 
if this campaign really took place in year 20 or 21, Shoshenq's reign 
would have begun in 946 or 945 BC. However, we can in no way 
be certain that the campaign took place immediately before the 
construction work started. 82 There is no reason why it could not have 

' 8 When dating according to the High Priests' years of office, we nevertheless have 
to consider the necessity of adding a few (possibly 2-3) years to Herihor's term of 
office under Ramesses XI subsequent to Payankh's term of office. Anyway, Ramesses 
XI's absolute length of reign (or the length of the whm-mswt-em) is uncertain. 

79 Kings I 14,25; II Chronicles 12,2. 

80 Cf Hornung, Untersuchungen, 24-29; Kitchen, TIP § 59; Beckerath, Chronologic 
68-70. This fixed point is only valid if we work on the assumption that the information 
concerning the kings' lengths of reign in the OT has been taken from reliable sources. 

81 R. A. Caminos, JEA 38 (1952), 46-61. 

82 Almost unanimous in the literature: presumably supported by the wish for at least 
one fixed point. 

DYNASTY 21 233 

taken place several years earlier. 83 In that case, the beginning of 
Shoshenq's reign would have to be set slightly later, and thus the entire 
Dyn. 21. 

83 Even if we could establish that there was a causal relationship between the cam- 
paign and the construction work, the work in Thebes could still have been begun long 
after the campaign. One could argue that the first priority was the enlargement and 
decoration of the LE temples and that simultaneous work in LE and UE was beyond 
the capacity of both the labour force and the architects. At least the temple of El- 
Hibeh in Middle Egypt had a depiction of the triumph, cf ASAE 2 (1901), 85-87; 
154-156; H. Ranke, Koptische Friedhqfe bei Kardra und der Amontempel Scheschonks I. bei el 
Hibe (Berlin & Leipzig, 1926), 50-52; pis. 19-21; E. Feucht, SAK 9 (1981), 105-117; 
pi. 2. 


Karl Jansen- Winkeln 

The prevailing conditions and patterns of rule during Dyns. 22— 23 were 
basically similar to the state of Egypt during Dyn. 2 1 . UE — with the 
important centers Thebes and Herakleopolis — was administered by a 
military governor who was simultaneously the High Priest of Amun; 
LE was directly governed by the king with residences in Memphis and 
Tanis (and in Bubastis as well, since Osorkon I). With the aid of their 
sons, the first kings of Dyn. 22 maintained their rule over the entire 
country. However, since the reign of Osorkon II at the latest, they 
gradually lost out to the powers of decentralisation, when (due to a 
divided inheritance?) clearly defined and separate spheres of power and 
local potentates appeared, particularly in LE. 1 In the same fashion, the 
separation of UE and LE remains tangible under Libyan rule. 2 

The most important chronological sources for UE are the records 
of the Nile levels, 3 the annals of the priests at Karnak, 4 the "Chronicle 
of Prince Osorkon",' and the statues (and other objects) belonging to 
dignitaries from certain families which permit detailed and extensive 
genealogies; 6 for LE, we only have the donation stelae' and the stelae 
from the Serapeum. 8 Altogether, there are relatively few actual dates 

1 It is not clear whether this regionalisation only came into existence at this time, 
or whether it existed earlier, i.e., already perhaps in Dyn. 21, but only became clear 
in the sources at this time (the most important sources are the donation stelae, and 
these only become abundant from later Dyn. 22, being totally absent in Dyn. 21). It 
is probable that there were at least incipient developments in this direction, which 
became more strongly expressed later. 

2 In fact, this division led to different cursive scripts used in the administration: the 
"anormal" hieratic in UE, and "Demotic" in LE. 

3 PM IF, 21-22; J.v. Beckerath, JARCE 5 (1966), 43-55; G. Broekman, JEA 88 
(2002), 163-178. 

4 PMW, 108; G. Legrain, RT 22 (1900), 51-63; Kruchten, Annales. 

5 PMIV, 35-36; Reliefs III, pi. 16-22; Caminos, Chronicle. 

6 Cf. TIP, §§ 157-205; Bierbrier, LNK, passim. 

7 See Meeks, Donations. 

8 See CSSM; PM IIP, 780ff. 


surviving from this period. As a rule — in contrast to the NK 9 — we lack 
a continuous series (or even relatively complete chain) of dates for any 
given sovereign, and thus by no means can we confidently suggest that 
the highest known date for any reign reflects its actual length. Given 
this paucity of dates, the chronology of this era is imprecise and uncer- 
tain in many respects. 

The actual means of dating was presumably the same as that of the 
NK, 10 as is suggested by the dates from one Serapeum stela. 11 These 
affirm that an apis bull, born in year 28 of Shoshenq III, was intro- 
duced on 1 /II/ Akhet of the same year: if the year began on 1 /I/Akhet, 
the Apis would have been a month old at the most — and this is highly 
unlikely. 12 Furthermore, his predecessor was buried in the same year, 13 
and there are generally several months between the burial of the pre- 
vious Apis and the introduction of the new one. 14 It follows that the 
regnal year still began with the accession of the king; unfortunately, 
there are no surviving accession dates for the TIP. 

1 . The Rulers of Unified Egypt of Early Dyn. 22 

According to Manetho, following Africanus, Dyn. 22 consisted of 9 
kings from Bubastis who ruled for 120 years: Sesonchis (21 years), 
Osorthon (15), three others (25), Takelotis (13) and three more (42). 15 
The family tree in the Serapeum stela of Pasenhor from year 37 of 
Shoshenq V ( c >-hpr-R c y 6 includes a reference to a King Osorkon who 
ruled six generations earlier, whose father, grandfather and great- 
grandfather were kings named Takelot, Osorkon und Shoshenq, while 
their forefathers were not kings, but rather Libyan princes. The non- 
royal origins of the earliest named king, Shoshenq, the exact corre- 
spondence of the names of the kings with those listed by Manetho for 

9 Cf. KM, VIII, 70-84. 

10 Thus also Beckerath, Chronologie, 10. It is a priori probable that the MK concept 
of "predating" was among the anachronisms introduced during Dyns. 25-26. 

11 Louvre SIM 3697, cf. CSSM, 21-22; pi. VIII (no. 22). R. Krauss drew my atten- 
tion to the importance of these dates. 

12 Cf. E. Winter, Der Apiskult im Alten Agypten (Mainz, 1983), 18. 

13 Stela Louvre SIM 3749, CSSM 19-20; pi. VII (no. 21). 

14 See L. Depuydt, JJVES 54 (1995), 123; Kienitz, Geschichte, 155. 

15 According to Eusebius only 3 kings in 49 years, namely Sesonchosis (21), Osorthon 
(15) und Takelothis (13). 

16 Louvre SIM 2846, cf. CSSM 30-31; pi. X (no. 31). 


this dynasty, and the period of time separating Pasenhor (nine generations 
to year 37 of Shoshenq V) clearly reveal that these were the first kings 
of Dyn. 22. In addition, the grandparents of this oldest Shoshenq link 
him to Dyn. 21, as he is the nephew of the third to the last king of 
that dynasty, Osorkon (Osochor). 17 This gives a sequence of 4 kings, 
each pair being father and son, for the start of Dyn. 22: Shoshenq I 
(Hd-bpr-K), Osorkon I (Shm-hpr-R c ), Takelot I (Hd-hpr-R) and Osorkon 
II (# Bistt Wsr-m!H-R). m 

Although each king is the son of a former king, this does not nec- 
essarily mean that each son immediately followed his father in office. 
It is entirely possible that other sovereigns can be fitted into the sequence. 
According to Africanus, Manetho inserts three other kings, and the fol- 
lowing are candidates for this: 

a) On his own documents, and in the patronymic of his son (a priest 
of Amun named Osorkon), the HPA, Shoshenq, son of Osorkon I and 
grandson of Psusennes II is designated as HP and Generalissimo and 
not as king. 19 Only on the London statue BM 8 does he enclose his 
name (in the titulary of HP) in a cartouche, adding the epithet mrjj- 
Imn. 20 

b) The statue Cairo CG 41292 from Karnak 21 was re -inscribed by 
a king Shoshenq with the throne-name MF-fopr-R' Stp-n-R\ and to the 
benefit of his "begetter" (msj sw) Psusennes II. It is entirely possible that 
this is an otherwise completely unknown son of Psusennes II, 22 but it 
seems more reasonable to assume that this is the (earlier?) high priest 
and son of Osorkon I, 23 who could easily have designated himself as 
"begotten" by Psusennes, his grandfather. 24 

17 J. Yoyotte, BSFE 77-18 (1977), 39-54; cf. above Jansen-Winkeln, Chapter II. 9. 

18 The throne-names were not listed on the stela of Pasenhor. Assigning the kings 
with these throne-names to the first kings of the dynasty results from a (i.a.) compar- 
ison with the kings appearing in the family tree of the Theban Nakhtefmut family, cf. 
TIP, § 88. For the throne-name of Takelot I, cf. VA 3 (1987), 253-258; TIP 3 , XXII-XXIII. 

19 The catalogue of these monuments (all from UE): K. Jansen-Winkeln, "Historische 
Probleme der 3. Zwischenzeit", JEA 81 (1995), 145-146. 

20 S. PM IP, 289. 

21 G. Legrain, Statues et statuettes de mis et de particuliers, III (Cairo, 1914), 1-2; pi. 1; 
J.v. Beckerath, Orientalia 63 (1994), 84-87; K. Jansen-Winkeln (n. 19), 147-148; pi. XIII. 

22 G. Broekman, GM 176 (2000), 39-46, considers Shoshenq M'-hpr-R' to be a son 
of Psusennes II who was able to assert his claims to be the royal successor of his father 
in Thebes at least, while Shoshenq I was recognized in LE (and dates in Thebes fol- 
lowed his reign). 

21 Thus also Beckerath (n. 21), 86; N. Dautzenberg, GM 144 (1995), 21. 
24 As jtj and z} can mean "grandfather" and "grandson". 


c) A number of kings were subsequently interred in the antecham- 
ber of the tomb of Psusennes I at Tanis, including two anonymous 
mummies 2 '' and a Shoshenq Hql-hpr-K Stp-n-R c , 26 who was presumably 
already more than 50 years of age 27 and whose throne-name bore a 
form reminiscent of early Dyn. 22 (before Osorkon II), 2!i and the same 
applies to the iconographic details of his shabtis. 29 The interment also 
included a pectoral of the great chief of the Ma, Shoshenq A, and a 
bracelet of Shoshenq I 30 — and thus the same person before and after 
the accession. As the individuals interred in the royal tombs often bore 
objects belonging to their parents, 31 this king is probably a son of 
Shoshenq I. 32 The commonly assumed identification of this king with 
the (earlier) HP and son of Osorkon I 33 does not appear to be very 

d) A king Shoshenq with the throne-name Twt-hpr-R c is known from 
the sherd Louvre E. 3 1886 from Abydos, 34 and apparently also from a 
fragmentary relief from Tell Basta. 35 This is evidently a king of the 
entire country and not a minor UE king or a local ruler. The form 
of the throne-name implies that he too belongs near the start of Dyn. 22. 

25 Possibly Siamun and Psusennes II, cf. J. Yoyotte in Tanis: L'or des pharaons (Paris, 
1987), 48. 

26 Montet, Tanis II, 36-51 

27 D. E. Derry, ASAE 39 (1939), 549-551. 

28 TIP, § 93. 

29 G. Broekman, GM 181 (2001), 29-31. 

30 Montet (n. 26), 43-45 (219; 226/227); fig.13. 

31 TIP, § 93; K. Jansen-Winkeln, VA 3 (1987), 256-257; D. Aston, "Takeloth II 
A King of the Theban Twenty-Third Dynasty'?", JEA 75 (1989), 139-153, esp. 

32 In addition he also bore the ring of a Dd-Pth-jw.f-'nh (Montet [n. 26], 44, fig. 13; 
46 [228]), perhaps his brother: a prince und 2nd/ 3rd Prophet of Amun of this name 
was interred in the cachette of Deir el-Bahri in year 1 1 of Shoshenq I (G. Maspero, 
Les momies royales de Deir el-Bahari (Paris 1889), 572-574; GLR, III, 284, n.2). He was 
presumably a son of Shoshenq I. 

33 TIP, §§ 93-94; 452; most recendy with new arguments Broekman (n. 29), 27-37. 
Rather than identifying Shoshenq Hq'-hpr-R' with the son of Osorkon I and grandson 
of Psusennes II, and thus being obliged to reckon with yet another new and hitherto 
unknown son of Psusennes, it appears more reasonable to identify the grandson of 
Psusennes II with the donor of CG 42192 and to identify Schoschenk Hq'-hpr-R' as a 
son of Shoshenq I, based upon his grave goods. 

34 According to the reading of J. Yoyotte, cf. M.-A. Bonheme, BSFE 134 (1995), 53. 

35 E. Lange, GM 203 (2004), 65—72. The arrangement of the cartouches does not 
allow one to deduce a coregency of Twt-hpr-R' (= Psusennes II) and Shoshenq (I) as 
Dodson does (BES 14 [2000], 9-10). Aside from this, Osorkon I is thus far consid- 
ered to be the first sovereign of the TIP documented in Bubastis. 


The HP Shoshenq ("II") is presumably identical with Shoshenq A4> c - 
hpr-R\ but most certainly did not have an independent reign, but rather 
was responsible for UE during the reign of his father. Shoshenq Hq'- 
hpr-R may have ruled briefly after his father, if Shoshenq I was his 
father, or perhaps after his brother Osorkon I. He could thus have 
been one of the "three other kings" Manetho places between Osorkon 
(I) and Takelot (I). 36 The same applies to Shoshenq Twt-hpr-R c who 
should most probably be put between Osorkon I and Takelot I. In 
contrast to his father and his son, not one single royal monument is 
known for Takelot I; 3 ' his brothers in UE probably dated according to 
his reign (cf below), but they do not name him. This could indicate 
that his rule was undisputed. 

For the first part of Dyn. 22 we would thus have the following kings, 
and dates: 

1. Shoshenq I; documented years 2, 5, 6, 10, 11, 13, 21 38 

2. Osorkon I: regnal years [l]-4, 6, 10, 11, 12, 23, 33 39 

3. Shoshenq Hql-hpr-K; no dates 

4. Shoshenq Twt-hpr-R: no dates 

5. Takelot I: years: 9, 40 dubious 5, 8, 13/14, 14 (cf. below) 

6. Osorkon II: years 12, 16, 21, 22, 23, 41 29(?) 42 

For Shoshenq I, Manetho's 2 1 years appear to be possible, and a reign 
of 35 years is quite probable for Osorkon I. 43 Only a year 9 is certain 
for Takelot I. The Nile level records nos. 16—21 are generally assigned 
to his reign: nos. 16 (year 5) and 20-21 (years lost) belong to the HP 

36 Takelot II cannot be implied, as he was an UE sovereign, cf. below. 

37 Cf. also TIP, §§ 95; 270. 

38 The highest regnal year on the rock stela of Gebel Silsila, cf. JEA 38 (1952), 
pi. XIII. 

39 Of these, only the year 10 in lines 2-3 of the "stele de l'apanage" {^AS 35 [1897], 
14) and year 12 of the Nile level record no. 2 (Beckerath [n. 3], 49) are explicitly related 
to Osorkon. Regnal year 33 is on the mummy wrappings of a burial, which also had 
a "counterweight" bearing the name and throne-name of Osorkon I, cf. J. E. Quibell, 
The Ramesseum (London 1998), 10-11; pi. XVIII. 

40 G. Daressy, RT 18 (1896), 52-53, earlier ascribed to Takelot II, cf. now Aston 
(n. 31), 144; TIP 3 , XXIII. 

41 Serapeum stela Louvre SIM 3090, s. CSSM, 17; pi. VI (no. 18). 

42 Nile level record no. 14, cf. below. 

43 Were one to follow Manetho here, we would still be obliged to emend 15 years 
to 35. Aside from the 33 which should in all probability to assigned to Osorkon I (cf. 
above), there are further indications of a long reign, cf. TIP § 89. 


Iuwelot who was still a youth in year 10 of the reign of his father, 
Osorkon I. 44 The year 5 must therefore relate to a successor of Osor- 
kon I. 45 The records nos. 17—19 are from the HP Smendes III, doubtless 
the brother and successor of Iuwelot; 46 no. 17 is from year 8, no. 18 
from year 13 or 14. A block, presumably from the Serapeum, bears 
the names of Takelot I and the HP of Memphis, Merenptah; 47 Mariette 
noted that this was found together with a stela from a year 14. 48 This 
might be a stela in Alexandria dated to a year 14 (without a royal 
name), and originally came from the Serapeum, as the inscription sug- 
gests. 49 This would thus support Manetho's 13 (full) years for Takelot. 
His possible predecessors (see above) have not left many traces and 
assuredly did not reign for a long period. 50 Thus for Takelot and the 
others, 15 years is a reasonable suggestion.' 1 One can therefore adopt 
Kitchen's suggestion of 21 + 35 + 15 years for the first 3 to 5 kings 
of Dyn. 22. However, these dates should be viewed as the minimum 
to which a few more years might be added. 

The length of the reign of Osorkon II is a matter of debate, and 
Manetho cannot aid here. The highest date which can with certainly 
be assigned to his reign is year 23 (see above), linked to an Apis bur- 
ial, where his son, the Crown Prince and HP of Memphis, Shoshenq 
D apparently also took part.' 2 Shoshenq D will thus have died after 
that time, but apparently before his father, 33 and thus Kitchen assigned 
Osorkon II 24—25 full years, to allow a margin for these events.' 4 

44 Lines 2-3 of the "stele de l'apanage", cf. %AS 35 (1897), 14. 
" But certainly not to Osorkon II, whose Nile level records take a different form, 
cf. Broekman (n. 3), 171. 

46 These records have exacdy the same form as those of Iuwelot and differ from all 
others, cf. most recendy Broekman (n. 3), 164; 170-171. 

47 CSSM, 18; pi. VII (no. 19). 

48 Mariette's remarks are, however, rather doubtful, cf. n. 47. 

49 G. Daressy, ASAE 5 (1904), 121 [XXIV]. The stela Louvre SIM 2810 (CSSM, 
18-19; pi. VII [no. 20]) of a Dd-Pth-jw.f-'nh of a year 10 [+ X] (without royal name) 
dates to a later epoque, cf. A. Leahy, SAK 1 (1979), 149. 

50 If there was a conflict over the throne, it is conceivable, that some of them ruled 
parallel to Takelot. 

51 If there really was an Apis burial in year 14 of Takelot, and the Apis buried in 
year 23 of Osorkon II was the successor of this bull (which is, of course, uncertain) it 
would favour placing year 14 towards the end of the reign of Takelot, as 26 years are 
the longest documented life of an Apis bull. 

52 Can no longer be verified, cf. TIP, § 81, with n. 77; CM 207 (2005), 76, n. 16. 

53 This was generally assumed because he is also designated as Crown Prince (rp't 
wr tpj n hm.f) in his tomb. 

54 TIP, 8 87. 


This logic is no longer tenable since Shoshenq D did in fact outlive 
his father. In his undisturbed burial was a chain of 8 W^^-amulettes 
(Cairo JE 86786), and one of them bore the name of Shoshenq III 
(Wsr-m> c t-R c Stp-n-Jmn Mrjj-Jmn z? Blstt Smq)P 

On the other hand, Aston has produced arguments that Osorkon 
IPs reign was clearly longer than previously assumed, and perhaps even 
40-45 years. 56 Aston's argument is based on the family trees of two 
Theban families which reveal that several generations lived in the reign 
of Osorkon II; other genealogical data likewise allegedly favours a longer 
reign; furthermore, there would be a whole series of HPA belong- 
ing to the reign of Osorkon II, and his three known sons would all 
have predeceased him. Of these arguments, only the family tree of the 
Nakhtefmut family" is really reliable, but this actually supports a 
relatively long reign for Osorkon II. Whether the genealogy of the 
Nebneteru-family 58 must also be understood in this sense is more debat- 
able: the statue Cairo CG 42225 was erected after the death of its 
owner, so that the name of the king and the high priest there could 
relate to the date of erection and not necessarily hint at the lifetime 
or term of office of the statue's owner. The other genealogical data 
which Aston introduces does favour a long life, but not necessarily a 
long reign for Osorkon II. As HPA under Osorkon II only his son 
Nimlot C, his grandson Takelot F 59 and Harsiese B are documented. 50 
Of the sons of Osorkon II, Harnakht C died as a child, Shoshenq D 
probably did outlive his father (see above), and thus effectively only 
Nimlot C predeceased him. 61 Nevertheless, I consider the basic sense of 
Aston's arguments to be correct. There is a Nile level record (no. 14) 

55 K. Jansen-Winkeln, "Der Prinz unci Hohepriester Schoschenk (D)", GM 207 (2005), 
77-78. It is conceivable (although rather improbable) that Osorkon II died immedi- 
ately after his son, and thus his successor may have been able to arrange for a gift 
for the burial. In this case, it would be certain that Shoshenq III was the immediate 
successor of Osorkon II (cf. below). 

56 Aston (n. 31), 145-148. 

57 Ibidem, 145. 

58 Ibidem, 146. 

59 Presumably the later Takelot II, cf. K. Jansen-Winkeln (n. 19), 138—139; Dautzenberg 
(n. 23), 24. 

60 Jansen-Winkeln (n. 19), 135-139. 

61 It is interesting to note incidentally that this HPA did leave hardly any traces in 
Thebes, being almost exclusively recorded in the genealogical records of his descen- 
dents there, and even there he is consistently designated as HPA and General of 


from the year 29 of an Wsr-mSH-R', who is most probably Osorkon II 
and not Shoshenq III or Osorkon III. 62 At the very least, the family 
tree of the Nakhtefmut-family clearly supports a reign for Osorkon II 
of more than the 24 or 25 years Kitchen allows him. In addition, it 
must be recalled that it is precisely from the reign of Osorkon II that 
we have comparatively numerous monuments, both royal and private: 
far more than from the eras of Shoshenq I, Osorkon I and Shoshenq 
III who are otherwise the best documented of the TIP. It is therefore 
not too bold to suggest a reign of at least 30-40 years for Osorkon 
II. In this era, it is hardly surprising that we do not have any dates 
from the final decade of the reign. 

The king Harsiese (A) also belongs to the period of Osorkon II: on 
the stelaphoric statue Cairo GG 42208 we see the complete titulary of 
Osorkon II, but the statue was dedicated "by the grace" of Harsiese. 63 
This Harsiese is known only from UE, 64 and was buried in Thebes. 
There are no known regnal years relating to him, and dating in his 
era presumably followed Osorkon II. 65 His reign should probably be 
assigned to the beginning of the reign of Osorkon II; 66 in any case, it 
is not chronologically relevant. 

The period from Shoshenq I to Osorkon II should have lasted about 
100-111 years (21 + 35 + 15 + 30-40), and would be ca. 945/40- 

62 Cf. Broekman (n. 3), 174-5. 

63 K. Jansen-Winkeln, Agyptische Biographien der 22. und 23. Dynastie (Wiesbaden: AUAT 
8, 1985), 453. 

64 Jansen-Winkeln (n. 19), 133—5. He is only documented as king, contrary to the 
common view, he is not documented as HPA even one single time. Earlier, he was 
viewed as the son of the HPA Shoshenq (II); but since it became evident that this 
was based on a mistaken reading (ibidem, 129-132), he has become an orphan. In 
the necropolis of the TIP at Herakleopolis was the burial of a woman named T'-nt- 
Jmn, in Tomb 4. According to the inscriptions of the tomb and grave goods (M. Perez- 
Die/P. Vernus, Excavaciones en Ehnasya el Medina (Madrid 1992), 50-59; 128-132; 
156—159; Docs. 21-26), she was wrt hnrt n Hrj-s.f, her father was the hm-ntr tpj (n) Jmn 
mr ml' fb'wtj Ns-bi 'nb-Ddt, and her mother was [>st-]>hbjt or Jhy (shortened version) and 
she is designated as mwt ntr. Represented together with Tj-nt-Jmn was a man named 
Osorkon, who was wr ',' n <pr> -Shm-hpr-R' . It necessarily follows that the HP Smendes, 
the father of T'-nt-Jmn, cannot have been Smendes II of Dyn. 21. If this is not an 
HP Smendes unknown from other sources, the only candidate is Smendes III of Dyn. 
22. As his wife is designated a "king's mother" (mwt ntr cannot be a sacerdotal title 
here), Smendes III must have had a son who became king, and who belongs to the 
generation of Osorkon II. Harsiese is the obvious candidate. 

65 However, the lack of dates could simply be the result of the type of documents 
which are preserved. 

66 Jansen-Winkeln (n. 19), 135. 


2. Takelot II 

Related to the length of the reign of Osorkon II and equally contro- 
versial is the question of the identity of his successor; the stela of 
Pasenhor has nothing to say on the matter. The HP Osorkon (B) who 
left a long inscription ("The Chronicle of Prince Osorkon") was a son 
of Takelot II (throne-name Hd-hpr-R c as with Takelot I), his mother was 
a daughter of the HP Nimlot (C) and a granddaughter of Osorkon II. 
In the inscription, the donations are at first dated according to the 
reign of Takelot II (until year 24), and then according to the reign of 
Shoshenq III (years 22—29), and thus a sequence of Osorkon II — Takelot 
II — Shoshenq III was deduced. 67 

D. Aston has dismissed this long established chronology for several 
reasons: 68 (1) Takelot II is only known in UE; (2) he has the epithet 
ntr hq> W>st in his throne-name; (3) his consort and children do not 
reveal any known links to LE either; (4) the genealogical details of his 
dependents hint that he belonged to the generation of the grandchil- 
dren of Osorkon II; (5) in the "Chronicle of Prince Osorkon", the years 
22—29 of Shoshenq III follow years 1 1—24 of Takelot II: were Takelot 
the predecessor of Shoshenq III, we would face a lacuna of more than 
two decades. Aston thus assumes that Takelot was a "Theban" ruler 
whose realm was restricted to UE, and thus that he ruled parallel to 
a LE sovereign. This would have major chronological consequences. 

K. A. Kitchen has strongly rejected this approach by attempting to 
disprove or disarm Aston's arguments: 69 Takelot II left relatively few 
traces in Thebes; other kings who definitely lived in a Delta residence 
had relations with Thebes; the epithet ntr hq> W>st was also borne by 
Shoshenq V (in Tanis); other kings of the TIP, such as Osochor, 
Psusennes II or Osorkon IV were rarely or not attested in LE, although 
they actually resided there. On the other hand, he suggests that the 
scenario leading to this "Theban" Takelot II is historically excluded: 
the Thebans would hardly have accepted a king in Thebes but rejected 

67 TIP, § 86; as noted already in principle by R. Lepsius (titer die XXII dgyptische 
Konigsdynastie, Berlin 1856, 271-274), who inserted yet another Shoshenq ("II", our 
Shoshenq D) between Osorkon II and Takelot II. 

68 Aston (n. 31), 140-144. 

69 TIP\ XXIII-XXIV; JEA 85 (1999), 247; BiOr 58 (2001), 383. 


and opposed his son as HP, and they would never have tolerated this 
HP as the later king Osorkon III. This argumentation is not convinc- 
ing. Takelot II and his son definitely belonged to a common "party" 
in the civil war; had Osorkon B been expelled from Thebes, the same 
would be true of his father. And it is rather doubtful that the opinion 
of the people (the "Thebans") would have had any role to play. Kitchen's 
replique does not dispose of the really decisive point: Takelot II and 
his entire family are attested only in UE and not at all in the Delta, 
and this point cannot be dismissed by references to such ephemeral 
rulers as Osochor, Psusennes II or Osorkon IV. The period from 
Osorkon II to Shoshenq III is the best documented of the TIP and 
both kings are demonstrably present in LE. That anyone else reigned 
in the same place for a quarter of a century, of whom (and whose 
dependents) no trace can be found, must be excluded. The genealog- 
ical connections of Takelot II and the sequence of years in the "Chronicle 
of Prince Osorkon" are likewise very clear. In addition, the HP Osorkon 
B disappears at the very moment (year 39 of Shoshenq III) when an 
otherwise unknown Osorkon appears as a new king; this is the only 
sovereign of Dyn. 22 who occasionally uses the title of HP in his royal 
name,' and his mother has the same name as the mother of the HP. 
It therefore follows that Osorkon B and Osorkon III are the same per- 
son, and that also demands that Takelot II must be placed parallel 
with Shoshenq III. There is thus a whole set of reasons supporting 
Aston's assumption, and nothing which contradicts it. Therefore, I con- 
sider the point to be certain. 

3. The LE Sovereigns of Dyn. 22 to Shoshenq V 

This would thus mean that Shoshenq III was the immediate successor 
of Osorkon II, and there is not the slightest hint of any other hitherto 
unknown king between them. 71 With Shoshenq III and his successors 
until Shoshenq V, we stand on firmer ground chronologically. For 

70 The Paleological Association of Japan, Akoris. Report of the Excavations at Akoris 
in Middle Egypt 1981-92 (Kyoto, 1995), 301-305; pi. 116; idem, Preliminary Report. 
Second Season of the Excavations at the Site of Akoris, Egypt 1982 (Kyoto, 1983), 
14-15; pi. 11. No other HPA is known from the period before Osorkon III with this 
name, aside from Osorkon B. 

71 Cf also Aston (n. 31), 144. 


Shoshenq III, recorded years include: 3, 5(?), 6, 12, 14, 15, 18(?), 22, 
23, 24, 26, 28, 29, 30, 31, 32, 33, 38, and 39. 72 An Apis-bull was 
buried in his year 28, and a stela commemorating the event' 3 was 
erected for the great chief of the Ma and HP of Memphis' 4 P>-dj->st, 
who was the grandson (through his mother Tz~B? stt-pri) and at he same 
time the great-grandson (through his father Vater Tkrjt) of Osorkon II. 
The successor of this Apis bull (introduced in the same year year, 
I /II/ Akhet) in turn died in year 2 (11/ Peret) of Pami, after reaching the 
age of 26 years. 73 Year 2 of Pami thus lies 26 years after year 28 of 
Shoshenq III. Were Pami the successor of Shoshenq III, the latter 
would have had a reign of no less than 52 years. In fact, however, it 
would appear highly probable that another king Shoshenq with the 
throne name Hd-hpr-K should be inserted here, 76 who was buried in 
the tomb of his predecessor. The most important piece of evidence 
here is a donation stela of year 10 from a King Shoshenq Hd-hpr-K, 11 
mentioning a Great Prince of the Libu named Niumateped, and a man 
apparently bearing the same name and title is documented from year 
8 of Shoshenq V.' 8 If, as would appear reasonable, this is the same 
person, then a king Shoshenq Hd-hpr-R' should be placed here, who 
reigned not long before Shoshenq V, but after Shoshenq III. As Shoshenq 
V probably reigned immediately after or following a very short inter- 
val after his father Pami, yet 13 years lay between year 39 of Shoshenq 
III and year 2 of Pami, for which we have no dates for Shoshenq III, 
then everything favours placing a 10—13 year reign of this Shoshenq 
Hd-hpr-K into this period. 79 The precise length of his reign is chrono- 
logically not very important since the total for the period between year 

72 Nile level record no. 22, cf. Beckerath (n. 3), 51; Annals of the priests, no. 7, cf. 
Kruchten, Annales, pi. 4; 19. 

73 Louvre SIM 3749, cf. CSSM, 19-20; pi. VII (no. 21). 

' 4 His son F?j.f-t>w-(m-)'wj-B'stt likewise bears the title of HP of Memphis on this 

75 Louvre SIM 3697, CSSM 21-22; pi. VIII (no. 22); cf. also the Stelae Louvre 
SIM 3736 and 4205, ibidem, 22-24; pi. VIII-IX (nos. 23/24). 

76 A. Dodson, GM 137 (1993), 53-58; TIF', XXV-XXVI. 

77 Meeks, Donations, 666 (22.1.10). 

' 8 While in year 3 1 of Shoshenq III, yet another great chief by the name Jnj-Jmn- 
n'j.f-nbw is documented, cf. J. Yoyotte, in: Melanges Maspero I, Orient ancien 4 (Cairo 
1961), 143 (§ 31). 

79 Numbered variously in the literature: lb, Ilia, IV or "quartus"; Ilia would be 
preferable, as this would eliminate all possible sources of misunderstanding. 


28 of Shoshenq III and year 2 of Pami is certain. For this king Pami, 80 
the years 2, 4, 5, and 6 are documented; from the structure of the text 
on his "annals" in Heliopolis, the presence of the years 3 and 7 can 
be deduced. 81 Were these "annals" to have covered the entire reign of 
Pami, 82 this would confirm Kitchen's assessment of 6 full years for the 
reign. 83 This assumption of a rather short reign for Pami is supported 
by the paucity of monuments he has left, and further by the fact that 
the reign of his son was quite long. However, the assumption of a mere 
6—7 years is not really certain. His son Shoshenq V followed Pami, 
probably as his immediate successor: a stela from the Serapeum from 
year 37 of Shoshenq V bears the name of the same (still living) donor 
as in year 2 of Pami. 84 It is thus improbable that this long period can 
be stretched any further. But, it cannot be excluded that another king 
(e.g., an older son of Pami) may have ruled between Pami and Shoshenq 
V, but then if at all, only very briefly. 85 

For Shoshenq V, the years 7, 8, 11, 15, 17, 19, 22, 36, 37, and 38 
are documented, 86 and the interval between year 28 of Shoshenq III 
and year 2 of Pami is 26 years long. If 6 full years are assigned to 
Pami, and Shoshenq V was his immediate successor, the period from 
Shoshenq III to year 38 of Shoshenq V would be 27/28 + 26 + 4/5 + 
37 years, and thus 94-96 years depending upon exactly when that Apis 
which died under Pami was introduced under Shoshenq III and when 
it died under Pami. The interval is probably 95 years. 

4. The Successors of Shoshenq V 

Shoshenq V is not among the rulers named on Piye's victory stela. He 
was probably already dead at the time. Appearing on a dedicatory stela 

80 See J. Yoyotte, RdE 39 (1988), 160-169. 

81 S. Bickel, L. Gabolde & P. Tallet, BIFAO 98 (1998), 31-56, esp. 41. 

82 Cf. ibidem, 42. 

83 TIP, § 83. 

84 Louvre SIM 3441 and 3091, cf. CSSM, 24-25; pi. IX (no. 25); 41; pi. XIII (no. 
42), cf. TIP, § 84, n. 97. 

85 However, the documented lifetime of the Apis-bulls do allow a somewhat longer 
period between Pami and Shoshenq V. A bull was buried in year 1 1 of Shoshenq V: 
between this one and the last known predecessor, buried in year 2 of Pami, are only 
15-16 years if Shoshenq V immediately followed Pami. 

86 PM IIP, 787-789. 


of his year 36 87 is Tefnakhte, the Great Chief of the Ma, commander 
and prince of the Libu, and again on another of year 38, that same 
Tefnakhte is called "Great Prince of the entire land". K!i The extension 
of this prince's power, which later obliged Piye to intervene, was thus 
already apparent at this time. It thus follows that the interval between 
the last years of Shoshenq V and Piye's campaign was not long. Shoshenq 
V is documented in Memphis and in diverse areas of the Delta, includ- 
ing Tanis, Bubastis, Buto and Kom Firin. On Piye's stela, Tefnakhte is 
lord of Memphis, Buto and Kom Firin; Iuput II rules in Leontopolis, 89 
Osorkon IV in Bubastis and the region of Tanis. 90 Osorkon IV would 
thus be spatially and temporally the successor of Shoshenq V, and the 
contemporary documents do not provide any reason to assign him to 
another dynasty. 91 

On the issue of the identity of Shoshenq's immediate successors, the 
temporal and spatial position of Manetho's Dyn. 23 could play a role. 
If Petubaste I and Osorkon III were UE rulers (cf below, section 5), 
then Manetho certainly did not take them into consideration. Thus 
they could not be those kings whom he assigned to his Dyn. 23 of 
Tanis (consisting of Petubaste, Osorkon, "Psammus" and "Zet"). Priese 92 
thus suggested that Osorkon rV (rather than III) be assigned to Manetho's 
Dyn. 23, A. Leahy has further elaborated on this idea. 95 Thus, Osorkon 
IV would be the successor of the ephemeral Petubaste, Shtp-jb(-n)-R c , 

87 From Buto (former collection Farouk), cf. Yoyotte (n. 78), 153, § 48; Meeks, 
Donations, 670 (22.10.36). 

88 From Tell Farain (in the storeroom?), cf. S. Sauneron, BSFE 24 (1957), 51; 53-54, 
figs. 1-2. The cartouches were left blank, but it unquestionably concerns the year 38 
of Shoshenq V, cf. TIP, § 84. The king is omitted on the other stela as well, which 
only has the year. 

89 A regnal year 21 is documented (J.-L. Chappaz, Genava 30 [1982], 71-81), but 
neither precedessors nor successors are known and thus he cannot be linked to any 

so jf R'-yifj, i s t De understood in this way, cf. Yoyotte (n. 78), 129, n. 2; F. Gomaa, 
Die libyschen Furstentumer des Deltas (Wiesbaden: Beihefte TAVO B, 6, 1974), 132-134. 
If not, then it means that, astonishingly, Tanis — one of the most prominent cities of 
the TIP — was not mentioned on the stela of Piye. This could only be explained if the 
rulers of Tanis declined to submit to the Nubian king. 

91 Leahy's interpretation, that "there is nothing to warrant his inclusion in . . . 
Manetho's Twenty-second Dynasty" (Libya, 189) is thus not entirely convincing. Inci- 
dentally, Osorkon IV is only documented with certainty on the stela of Piye: the other 
references could also be assigned to Osochor of Dyn. 21, cf. Leahy, Libya, 189; F. 
Payraudeau, GM 178 (2000), 75-80. 

92 ZAS 98 (1972), 20, n. 23. 

93 Libya, 186ff. 


who is known from Memphis and Tanis (among other places), 94 and 
otherwise identified with Putubisti of the annals of Assurbanipal. 9 ' 1 Aston 96 
and Beckerath 97 have both followed him. At the very least, this would 
be a means of integrating Manetho's Dyn. 23 into the previously known, 
although identifying Petubaste Shtp-jb-R with the Putubisti of the Assyrians 
is at least equally plausible. In any case, the result would be that 
Manetho's Dyn. 23 would be nothing but a continuation of Dyn. 22. 98 
As regnal years have not been preserved from the reign of either 
Osorkon IV, nor of his supposed predecessor, Petubaste Shtp-jb-R, and 
the transition from Shoshenq V ( — Petubaste) — Osorkon IV is to be 
dated to before Piye's campaign, this possible insertion of a Petubaste 
(Manethonis gratia) is not of chronological significance. Osorkon IV is 
only dated through the campaign of Piye. Were he the king Shilkanni 
who paid tribute to Sargon II (cf. below), then he will still have been 
in office around 715/716. 

5. UE Kings and Dynasties from Takelot II to Dyn. 25 

Along with two Lower Egyptian rulers, the stela of Piye names two 
Upper Egyptians: Nimlot D of Hermopolis and Peftjau'awybast of 
Herakleopolis. At this time, Thebes itself will have already been under 
Nubian control, but before this time we find Harsiese A and Takelot 
II (cf. above, section 2) as UE kings who ruled Thebes. Of Kitchen's 
Dyn. 23 (Petubaste I, Iuput I, Shoshenq IV, Osorkon III, Takelot III, 
Rudamun und Iuput II, as well as perhaps also Shoshenq VI; 99 Residence: 

94 Cf. Habachi, %AS 93 (1966), 69-74; pis. V VI; P. Montet, Le lac sacre de Tanis 
(Paris, 1966), 63-5; pi. XXX. 

95 Cf. TIP, § 357. 

96 Aston (n. 31), 140. 

97 Chronologic, 99. 

98 This could have been another branch of the family, with deeper roots in Tanis 
than Bubastis. In any case, according to our present knowledge, Manetho's king list 
of Dyn. 23 is more or less useless for the historical (and chronological) reconstruction: 
the last two of his four kings are virtual phantoms, the first two cannot be identified 
with certainty, and the note that the first Olympiad took place during the reign of 
Petubaste is generally dismissed as a later invention, calculated by the Christian chrono- 
graphers who used Manetho, cf. TIP, § 419, n. 134; Redford, King-lists, 311-312; 
Beckerath, GM 147 (1995), 9. 

99 Shoshenq VI [Wls-ntr-R, cf. TIP, §§ 67; 110; 146; 336; M.-A. Bonheme, Pes noms 
royaux de I'Egypte de la Troisi'eme Periode Intermediaire [Cairo: BdE 98, 1987], 140-141) is 
not considered in the following, since his very existence is debatable, and there is in 
any case no indication of where he should be placed chronologically. 


Leontopolis) 100 Iuput II is only documented in LE, Petubaste I mainly 
in UE, but a few times in LE; the others are known exclusively from 
Upper and Middle Egypt. Osorkon III is the father of Takelot III and 
Rudamun, and the later is the father-in-law of Peftjau'awybast. All of 
the members of this family are known exclusively from UE sources. 101 
They are doubtless UE rulers in the tradition of Harsiese A and Takelot 
II, and thus are not Manetho's Dyn. 23. The issue is thus the temporal 
relationship of those kings known from UE sources to one another and 
to the kings of Dyn. 22. The sources allow for the following synchronisms: 

a) In the "Chronicle of Prince Osorkon", years 22—29 of Shoshenq III 
follow year 24 of Takelot II. 102 This suggests that Takelot II became 
king in UE during the reign of Osorkon II (as Harsiese A before him, 
but with his own count of regnal years) and that in his year 4, 
Shoshenq III became the successor of Osorkon II (in LE). 

b) The year 1 2 of a king who can only be Shoshenq III corresponds 
to the year 5 of Petubaste I, with Harsiese (B) as HPA. 103 Petubaste I 
thus began his reign in year 8 of Shoshenq III (= year 1 1 of Takelot II) 
and HP Harsiese (B) is linked to this regency. Harsiese (B) is subse- 
quently documented in the years 18 and 19 of Petubaste (= years 25 
and 26 of Shoshenq III), 104 and previously in year 6 of Shoshenq III, 1(b 
and already under Osorkon II. 100 A Takelot (E) was HP at the latest 
from year 23 of Petubaste, 10 ' who then assumes Harsiese's post. 

It is therefore highly probable that the "rebellion" of year 1 1 of 
Takelot II mentioned in the "Chronicle of Prince Osorkon" was the 
accession to the throne of Petubastis, 108 which was understood as a 
usurpation, as he thus became a kind of rival king to Takelot II. The 

100 TIP, §§ 102; 297; 519; p. 588. 

101 And the same applies, as described above, to Takelot II, the father of Osorkon III. 

102 Reliefs, III, pi. 22, Z.7-22. 

103 Nile level record no. 24, cf. Beckerath (n. 3), 51. On the identification of the 
unnamed king as Shoshenq III, cf. TIP, §§ 106-107. On purely technical grounds, fol- 
lowing the chronology proposed by Aston, Takelot II could also be considered, but 
historically, he is out of the question, as an opponent of Petubaste and Harsiese B. 

104 Nile level record nos. 28 and 27, cf. Beckerath (n. 3), 52. 
10 ' Nile level record no. 23, cf. Beckerath (n. 3), 51. 

106 On the statue, Cairo CG 42225, for this, cf. Jansen-Winkeln (n. 19), 135-6. 

107 Nile level record no. 29, cf. Beckerath (n. 3), 52. 

108 Chronicle of Prince Osorkon, A, 22ff., cf. Reliefs, III, pi. 16; 18. 


HP Osorkon B is documented for years 1 1 and 12 in Thebes, 109 whereas 
Petubaste I and Harsiese B are not, but another revolt erupts in year 
15 of Takelot, 110 and exactly in this year, Petubaste and Harsiese B 
reappear in the Theban sources. 111 In year 24 and 25 of Takelot (= 14 
and 15 of Petubaste I), Osorkon B donated offerings in Thebes, and 
at this time Petubaste and Harsiese are not documented here. Evidently, 
there were two parties in this civil war: Osorkon B and his father 
Takelot II on the one hand, and Petubaste I and the HP Harsiese B, 
later Takelot E, on the other. 112 This Takelot is also mentioned in the 
year 6 of a king Shoshenq Wsr-mUH-R' Mrjj-Jmn, ni who cannot be 
Shoshenq III, 114 but must rather be an another (certainly UE) King 
Wsr-m> c t-R c Shoshenq (IV). 115 

c) The highest documented regnal year for Takelot II is year 25, 116 
and as in the donation lists of the "Chronicle of Prince Osorkon", year 
24 of Takelot II is followed by year 22 of Shoshenq III, it was appar- 
ently his last. 117 Despite publicly announced claims, 118 the successor of 
Takelot II was not his son Osorkon B: the latter is still General and 
HP in year 39 of Shoshenq III. 119 It was presumably Iuput I who was 

109 The son of Takelot II who commissioned the Chronicle of Prince Osorkon. 

110 Chronicle of Prince Osorkon, B,7, s. Reliefs, III, pi. 21. 

111 Nile level record no. 24, cf Beckerath (n. 3), 51. 

112 The role played by Shoshenq III in these events is not evident. 

113 Nile level record no. 25, cf. Beckerath (n. 3), 52. 

114 Cf. Aston (n. 31), 151: Shoshenq III does not use the epithet Mrjj-Jmn in his 
throne-name, and there is already a Nile level record (no. 23) for his year 6, naming 
HPA Harsiese. 

115 The latest documented date for him is year 6, cf. Beckerath (n. 3), 52. (Nile level 
record no. 25); Jacquet-Gordon, Graffiti, 40-41 (no. 100). 

116 Donation stela Cairo JE 36159, cf. ASAE 4 (1903), 183. 

"' The years 24 and 26 (without the king's name, cf. Capart, BMRAH, 3. serie, 
13, 1941, 26), are recorded on the mummy wrapping Brussels E.7047b/c of a mrjj-ntr 
named JVs-p> -ntr-n-R' var. Ns-ntr-ft-R'. As the father of this man is Ns-r-Jmn (Cartonage 
Berlin 30, cf. AIB II, 381-382), Kitchen {TIP § 86, n.115; 294) and Bierbrier (LNK, 
71) have both identified him as Ns-ft'-R', son of Ns-r-Jmn (I), the donor of the statue 
Cairo CG 42221, whose family tree (TIP, § 166) suggests that he belongs roughly in 
the period of Takelot II, and they have thus deduced a year 26 of Takelot II. Since, 
however, both the name (Ns-pl-ntr-n-pS-R' vs. Ns-pl-R', cf. M. Thirion, RdE 46 [1995], 
181-182) and the tide (mrjj-ntr vs. hm-ntr n Jmn-R' njswt ntrwjmj->bdfn pr Jmn hr z? tpj) 
of these individuals differ, this identification (and thus a possible source for a year 26 
of Takelot II) cannot be maintained. 

118 Chronicle of Prince Osorkon, A, 53, cf. Reliefs, III, pi. 16; 17; CPO, §§ 101-102. 

119 Nile level record Karnak no. 22, cf. Beckerath (n. 3), 51; Annals of the Priests 
at Karnak, no. 7, cf. Legrain (n. 4), 55-56; Kruchten, Annales, pi. 4; 19. 


the successor, for year 16 of Petubaste I corresponds to year 2 of a 
king Iuput (I), 120 and thus his year 1 (corresponding to year 15 of 
Petubaste and year 22 of Shoshenq III) follows immediately on the last 
full year of Takelot II. As these dates match, it is more probable that 
Iuput I was the successor of Takelot, and not a "short-lived coregent" 
of Petubaste. 121 On the other hand, Shoshenq IV may have been the 
successor of Petubaste as Petubaste appears initially together more fre- 
quendy with the HP Harsiese B, and then with Takelot E, who him- 
self is then named likewise together with Shoshenq IV (cf. above). These 
synchronisms produce the relations presented in Fig. II. 10. 1. 122 

King Petubaste is documented in Thebes with the throne-name Wsr- 
mS c t-R c Stp-n-Jmn and with the unique epithet z? >st. 123 A king with the 
same prenomen and throne-name, but with the epithet z? BSstt is known 
from a donation stela from Memphis (year 6), 124 from Herakleopolis or 
the eastern Delta (?), 125 and Bubastis (year 23), 126 as well as on a statue 
of uncertain provenance. 12 ' This has been interpreted as being two 
different kings with the same prenomen and throne-name, 128 but this 
is hardly plausible. 129 The idea that both the UE and LE Petubaste 
would have the same highest known date of 23 years appears rather 
suspicious. In addition, one of the Theban retainers of Petubaste, the 

120 Nile level record no. 26, cf. Beckerath (n. 3), 52. 

121 In TIP, § 448; cf. also Aston (n. 31), 151. Against this, one could argue that all 
of the other synchronisms in the Nile level records give only the links between the 
rulers of one "party" to the LE king (Shoshenq III). If Iuput I was the successor of 
Takelot II, he should have belonged to the foes of Petubaste. However, from the 
Chronicle of Prince Osorkon (B,7ff.) it is evident that at this time, there was a tem- 
porary unity among the various rivals in the civil war (cf. Jansen-Winkelen [n. 19], 
140-141 on this). 

122 Abbreviations: NLR, Nile Level Records, cf. Beckerath (n. 3), 43-55; OC = 
Caminos, Chronicle; OC, A = Reliefs, III, pi. 16-19; B = ibidem, pi. 21; C = ibidem, 
pi. 22; AP = Annals of the Priests at Karnak, cf. Legrain (n. 4), 51-63; Kruchten, 
Annates; Stela 22.8.26 = Meeks, Donations, 669 [22.8.26]. Years in brackets are postulated. 

123 Nile level record no. 24; Beckerath (n. 3), 51. 

124 Cairo JE 45530, cf. Schulman, JARCE 5 (1966), 33-41; pi. 13. 

125 Copenhagen Ny Carlsberg AElN 917, cf. O. Koefoed-Petersen, Recueil des inscrip- 
tions hieroglyphiques, pi. 5; J. Yoyotte, BIFAO 58 (1959), 97 (2); Meeks, Donations, 671 

126 Florence 7207, cf. R. A. Caminos, Centaurus 14 (1969), 42-46; pi. 1-2. 

127 Gulbenkian Museum Lisbon, cf. M. Hill, Royal Bronze Statuary from Ancient Egypt 
(Leiden/Boston, 2004), 155-156; pi. 18 (12). 

128 E.g., A. S. Schulman, JARCE 5 (1966), 37-39; Beckerath, GM 147 (1995), 9-13. 

129 Cf. B. Muhs, JEA 84 (1998), 223; J. v. Beckerath in: Es werde niedergelegt als Schriftstikk 
Festschrift fur Hartwig Altenmiiller zum 65. Geburtstag (Hamburg 2003), 31-36. 




II. 10.1 



Dyn. 23 (UE) 

Rival Kings 

High Priests 


Osorkon II 

Harsiese B 

CG 42225,a 

Takelot II 


henq III 





Harsiese B 

NLR, no. 23 

Petubaste I 




Osorkon B 

OC, A 18-53: 1st 




Osorkon B 

OC, B 1-6 




Osorkon B/ 
Harsiese B 

NLR, no. 24; 

OC, B 7: 

2nd "rebellion" 




AP, no. 1, 1.1 




Harsiese B 

AP, no. 1, 1.2; 
no. 2, 11.1/3 




Osorkon B 

OC, C 7 (cf. B 


Iuput I 



Osorkon B 

OC, C 12 (year 
22); ASAE 4, 




Osorkon B 

NLR, no. 26 

(y. 2/16); OC, 
C 12 




Osorkon B 

OC, C 13-16 




Osorkon B/ 
Harsiese B 

NLR no. 28; 
OC, C 1-2; 17: 
3rd "rebellion" 




Harsiese B 

NLR, no. 27; 
Stela 22.8.26 




Osorkon B 

OC, C 12; 17 




Osorkon B 

OC, C 22 




Takelot E 

NLR, no. 29 

Shoshenq IV 


Graffiti, 85 







Takelot E 

NLR, no. 25 



Osorkon B 

NLR, no. 22; PA, 

no. 7, 11.1-3 


Osorkon III 



prophet of Amun and royal scribe Hr (IX), is unexpectedly documented 
at Memphis, 130 and perhaps also in Tell el-Balamun. 131 There can only 
be one single king Petubaste, who used the epithet z? B>stt in LE. He 
may have been a rival king who attempted to re-establish a unified 
kingship over the entire land, a situation which had ceased to exist at 
the very latest by Takelot II. Regardless, the "dynasty" of Petubaste is 
not chronologically relevant. It is not known when Shoshenq IV suc- 
ceeded Petubaste, the length of whose reign is likewise unknown. This 
dynasty presumably ended in year 39 of Shoshenq III, at the latest. 132 
By contrast, the dynasty of Takelot II can be followed: a year 12 is 
documented for his presumed successor, Iuput I (cf above), 133 and his 
successor can only have been Osorkon B/III. He appears for the last 
time in year 39 of Shoshenq III, as High Priest. As he had this office 
since year 1 1 of Takelot II (= year 8 of Shoshenq III), and then reigned 
for 28 years as king, he must have become king in or shortly after 
year 39 of Shoshenq III. If he, as is probable, followed immediately 
after Iuput I, the latter must have reigned for at least 17 years. 

For Osorkon III, the regnal years 1(?), 3, 5, 6, 14(?), 15, x + 6, 23(?) 
and 28 are documented, with his regnal year 28 being equal to year 
5 of his son Takelot III, 134 the only completely unambiguous coregency 
in the TIP. 135 For Osorkon III, 23 full years can be accounted for, 
and for Takelot III, years 5, 6, and 7 are clearly attested. 136 If Osorkon 

130 K. Jansen-Winkeln, SAK 27 (1999), 123-139; pis. 1-4. 

131 A. J. Spencer, Excavations at Tell el-Balamun 1995—1998 (London 1999), 13—15; 
83-86; 90-91. 

132 In this year, the HPA Osorkon B claimed that he and his brother defeated all 
of those with whom they fought, cf. Legrain (n. 4), 55-56; Kruchten, Annates, pi. 4; 
19. It would still be conceivable that the later "dynasty" of Hermopolis (Nimlot D and 
Thotemhat) continued that of Petubaste, as Hermopolis could have been a major cen- 
ter in Petubaste's "rebellion", cf. Jansen-Winkeln (n. 19), 142. However, there does not 
appear to be any trace of a temporal link between these regents. 

133 Aside from the graffito of year 9 of Iuput, the same priest also left graffiti from 
years 9 and 12 (without the name of a king), cf. Jacquet-Gordon, Graffiti, 84-85 (nos. 

134 Nile level record no. 13; Beckerath (n. 3), 50. For the uncertain numbers, cf. 
Jacquet-Gordon, Graffiti, 41 (nos. 101: year 1); 68-9 (no. 190: year 14); 69 (no. 191: 
year 23). 

135 This conregency is also confirmed by the statue Cairo CG 42211, dated by the 
cartouches of njswt-bjt Mrjj-Jmn 3 >st Tkrt and z> K Mrjj-Jmn 3 >st Wsrkn, cf. Jansen- 
Winkeln (n. 63), 470. 

136 Daressy (n. 40), 51-52. 


III ascended the throne in year 39 of Shoshenq III, then the tempo- 
ral relationship between Dyn. 22 and the UE rulers of the line of 
Takelot II will have been that presented in Fig. III. 10.2. If he became 
king at a later date (year 40 or 41), then the dynasty must be pressed 
down a bit, but this can hardly be a matter of more than a few years. 137 
Of his successors, only his well documented son Takelot III spatially 
and temporally anchored in Thebes. The length of the reign remains 
unclear: he is occasionally assigned a reign of more than 6 full years, 
and not least because several of his children were still alive shortly 
before 700 as the family trees of their descendents and the style of 
their tombs reveal. 138 F. Payraudeau has recently attempted to link a 
year 14 of a Takelot z$ 1st in P. Berlin 3048 to Takelot III rather than 
Takelot II. 139 This is possible but by no means certain. 140 However long 
he reigned, the problem of the "generation shift" does not disappear: 
perhaps Takelot III and/or Osorkon III only became fathers late in 
their lives. 

At the very latest, after the reign of Takelot III the situation in UE 
becomes quite obscure. At the time of the Piye campaign, the Nubians 
ruled the Thebaid, while other UE kings were in Hermopolis and 
Herakleopolis. The later successors of Osorkon III were thus driven 
out of Thebes. There is no clear indication of when this happened, 
but at the very latest the inauguration of Amenirdis I as the adoptive 
daughter and heir of the Divine Wife Shepenupet I marks that Thebes 
was definitely governed by the Nubians. According to Kitchen, it was 
Piye, the brother of Amenirdis, who ordered the adoption, 141 but Morkot 

137 On the condition that the HPA Osorkon B and Osorkon III were in fact one 
and the same person, cf. above, section 3. 

138 Cf. Aston & Taylor, in: Leahy, Libya, 138-143. 

139 GM 198 (2004), 82-85. Palaeographically it appears probable that on the same 
papyrus (debt note) one should read year 23 (rather than 13), cf. S. Vleeming, OMRO 
61 (1980), 3, n. 14; B. Menu, CRIPEL 1 (1973), 89-90; K. Donker van Heel, in: 
K. Ryholt, ed., Acts of the Seventh International Conference of Demotic Studies (Copenhagen: 
CNI Publications 27, 2002), 142. 

140 The palaeography can hardly aid with the date as there are already very cur- 
sive texts in Dyn. 21, cf. M. Malinine, in: Textes et langages de I'Egypte pharaonique I (Cairo, 
1973), 31. The identification of some ancestors of the scribe with individuals from 
dated contexts is unreliable in the extreme, or would even favour a date under Takelot 
II (as with the vizier Hrj). Nor can an argument be made using the Overseers of the 
Treasury, as four of them appear in this one Papyrus (cf. Donker van Heel [n. 139], 

141 TIP, § 122. 



Fig. III. 10.2 (Abbreviations: see Fig. III. 10.1) 

Shoshenq III 

Takelot II 

Year 1 

Year 4 




luput I 




Osorkon III 




Shoshenq Ilia 





Shoshenq V 









24 = 
28 = 

1 Takelot III 


7 (sole rule) 

has convincingly shown that it was probably her father Kashta who 
installed her. 142 This would mean that the successors of Osorkon III 
were swiftly removed from Thebes. If Piye's campaign (in his year 20) 
took place within five years of the death of Shoshenq V (see below, 
section 7), then his reign must have begun at the latest in year 25 of 
Shoshenq V, and probably somewhat earlier. The inauguration of 
Amenirdis could thus have taken place in years 20—24 of Shoshenq V. 
As year 28 was probably the final year of Osorkon III, and corre- 
sponds to year 8 of Shoshenq V, at the earliest (cf above), his suc- 
cessors have a mere 10—15 year in Thebes, before they had to withdraw 
to the North. All of their dated sources from Thebes must be assigned 
to this short period. 

In: S. Wenig, "Studien zum antiken Sudan", Meroitica 15 (1999), 194-196. 


Aside from Takelot III, the following UE kings are known from the 
period after Osorkon III. 

• Rudamun, the brother of Takelot; 143 no known regnal years. 

• Peftjau'awybast, the son-in-law of Rudamun, king of Herakleopolis 
at the time of Piye's campaign; 144 regnal year 10 is documented. 145 

• G. Broekman has recently shown that it is highly probable that 
there was an UE king Shoshenq ("VII") with the epithet z? >st and 
the throne-name Hd-hpr-R Stp.n-R c , m who was recognized as king 
in Thebes in his regnal year 5, 14/ and who is to be inserted after 
Shoshenq III and thus also after Takelot III. 

• Another candidate would be the king Iny who is documented sev- 
eral times in Thebes (including a regnal year 5) and perhaps also 
in Abydos. 148 

• In addition, there is a dynasty residing in Hermopolis, whose most 
prominent member, Nimlot D, is chronologically anchored in the 
stela of Piye. His predecessor or (more probably) successor could 
have been Thotemhat, 149 and a later successor may have been 
Padinemti(?). 150 

143 Cf. O. Perdu, RdE 53 (2002), 157-178, for this person. 

144 Even if his power was restricted to the Herakleopolis region, during this period 
when the Nubians controlled the Thebaid and there appeared yet another UE king- 
dom, he could still have been the heir of an UE dynasty with a much larger realm. 
In Herakleopolis and the surrounding area at least, the dynasty of Takelot II is well 
documented, e.g, the HPA Osorkon B (cf. Caminos, Chronicle, §§ 28-30) and the later 
Takelot III {ASAE 37 [1937), 16-24). Payraudeau's ([n. 139], 79-81) attempt to dis- 
tinguish the general of Herakleopolis from the son of Osorkon III, who bears the same 
name, cannot be accepted in view of the fact that both are HPA and had a mother 
with the same rather uncommon name. 

145 Donation stelae Cairo JE 45948 and 11/9/21/14, cf. G. Daressy, ASAE 17 (1917), 
43-45; ASAE 21 (1921), 138-139. 

146 Broekman (n. 3), 163-78, esp. 176-177. 

14/ The only certain document is the Nile level record no. 3; Beckerath (n. 3), 49, 
hitherto assigned to Shoshenq I. However, one cannot exclude a possible reference to 
Shoshenq Ilia; his predecessor Shoshenq III is in fact mentioned in his last (or next 
to last) year in the Nile level records, cf. Broekman (n. 3), 176. It is conceivable that 
there was still resistance after Osorkon III ascended the throne, and that one of his 
enemies was able to establish himself briefly in Thebes, and dated according to the 
LE king. It is highly probable that the Nile level record no. 45 does not belong to 
Shoshenq VII (cf. Broekman [n. 3], 177); there does not remain any time for a year 
17/19/25 of a sovereign in Thebes before the Nubians after Osorkon III (cf. above). 

118 J. Yoyotte, CRIPEL 11 (1989), 113-131; pi. 14. 

149 For him, cf. H. Wild, RdE 24 (1972), 209-215; P. Spencer & A. Spencer, JEA 
72 (1986), 198-199; pi. 21; Bierbrier, LNK, 84. 

150 For him most recently, cf. A. Leahy, JEA 85 (1999), 230-232. 


As the brother of Takelot III, Rudamun was most probably his suc- 
cessor, as is generally assumed. It is, however, remarkable, that he is 
better documented in Hermopolis than in Thebes. 151 It is thus also con- 
ceivable that Rudamun became king in Hermopolis after the death of 
his father, alongside his brother Takelot in Thebes (and Herakleopolis?). 
The Libyan period does reveal a tendency to multiply both rulers and 
principalities. The line of Takelot would then have been reduced to 
Herakleopolis after the Nubian intervention. Shoshenq "VII" is only 
documented in Thebes, with a year 5. He too can belong only to the 
dynasty of Osorkon III (as a son of Takelot III?). If Rudamun was the 
successor of Takelot III (in Thebes), Shoshenq VII would most prob- 
ably have been a successor of Rudamun, although a sequence of 
Takelot — Shoshenq — Rudamun cannot be excluded. 1 ' 2 If Rudamun was 
a local ruler in Hermopolis, then Shoshenq VII would have followed 
immediately after Takelot. The year 5 of king Iny should be situated 
roughly two generations after year 4 of a king Shoshenq; 153 and this 
may have been Shoshenq III, IV or VII. Were it Shoshenq III, the 
reign of Iny would fall under the reign of Osorkon III, and that is 
improbable. Otherwise, he should be assigned either to the period after 
Takelot III (successor of Shoshenq VII?), or indeed placed in Dyn. 25. 
Unusually his name was effaced, and thus he might have been a pre- 
tender (during the reign of Osorkon III or Dyn. 25), 154 in which case 
the reign would be of no chronological relevance. 

In any case, the rulers of the house of Osorkon III were swiftly 
evicted from Thebes. The Peftjau'awybast of Herakleopolis named on 
the stela of Piye is the last of this line. The "dynasty" of Hermopolis 
(whether from Rudamun or by another line) may have been founded 
by descendents of Osorkon III, but it could equally easily have been 
the late revival of the rival dynasty of Petubaste. 

151 Cf. Perdu (n. 143), 169-170. 

152 Cf. G. Broekman, "The Chronological Position of King Shoshenq Mentioned in 
Nile Level Record No. 3 on the Quay Wall of the Great Temple of Amun at Karnak", 
SAK 33 (2005), 75-89. 

153 Graffito no. 1 1 from the roof of the temple of Khonsu, cf. H. Jacquet-Gordon, 
in: Hommages a la memoire de Serge Sauneron I (Cairo, 1979), 174-183; pi. 27-28; Yoyotte 
(n. 148), 115. 

154 Cf. Yoyotte (n. 148), 131. A "reign" of at least 4 years for a rival king could be 
possible, but it would be quite unusual, and particularly so in Dyn. 25. 


The chronology of the UE kings after Osorkon III thus remains quite 
uncertain: there are only a few (low) dates, and it is unclear which 
kings reigned in parallel and which in succession. Of the kings attested 
on the stela of Piye, Nimlot D cannot be linked to either a predeces- 
sor or a successor, 155 and Peftjau'awybast can only be identified genealog- 
ically as the son-in-law of Rudamun. 

The familiar "graffito" from Wadi Gasus could offer a chronologi- 
cal connection with the following Dyn. 25. 156 To the right is the car- 
touche of the Divine Adoratrice Amenirdis (I), above which is regnal 
year 12, to the left the cartouche of the Divine Wife Shepenupet (I), 
above which is regnal year 19, both names have the epithet "living" 
( c It is today agreed that the year 12 of Amenirdis can only be 
related to Piye, 15 ' and thus the year 19 should be assigned to one of 
the UE rulers recognized by Piye. It is thus immediately clear that 
Takelot III cannot possibly be the king designated by the year 19 of 
the graffito 1 '' 8 since his year 19 must have corresponded to year 22(~25, 
or so) of Shoshenq V (cf. above, Fig. III. 10. 2), and thus clearly before 
year 12 of Piye in whose year 20 the campaign to the North took 
place, Shoshenq V, who reigned at least 37 years, no longer in office. 

Rudamun would only be a candidate for the year 19 if he was not 
the predecessor of Shoshenq VII, as they were certainly not dating in 
Thebes according to the dynasty of Osorkon III 19 + 5 years after 
Takelot III (cf. above). The year 19 can also be linked to Shoshenq 
VII or Peftjau'awybast. In any case, it should certainly be someone rel- 
atively close to Piye's house: an ally. Nimlot D of Hermopolis would 
thus also be a candidate. He appears in an ambivalent fashion on the 
stela of Piye: on the one hand the Nubian king expresses his particu- 
lar irritation over the alliance with Tefnakhte of Sais, while on the 
other, he is given preferential treatment. 1 ' 9 This can be most easily 
explained by the fact that he was an ally of the Nubians who then 

155 ]\j re g na i year is preserved, and the same is true of several other members of 
this dynasty: Rudamun, Thotemhat, and Padinemti; for the latter two, not even the 
exact position in the sequence of the "dynasty" is known. 

156 L.-A. Christophe, BIE 35 (1952/53; 1954), 141-152. 

157 TIP, §§ 143-145. 

158 Thus Payraudeau (n. 139), 85-86. 

159 He is the only prince admitted into the Palace to Piye, cf. the great stela of vic- 
tory, 11. 148-53 (Urk. Ill, 54) and is the only one pictured standing, but actually like 
a woman, with a sistrum in his hand. 


switched sides. 160 As an ally before these events, he would have been 
a suitable candidate for the double dating, and in fact he does appear 
a second time with the Divine Wives Shepenupet I and Amenirdis I. 161 
Nimlot D thus appears to me to be a particularly suitable candidate 
for the year 19 in this graffito. Chronologically, however, this does not 
aid at all: in temporal terms, neither Nimlot D nor the other possible 
candidates can be pinned down to sufficiently narrows slots in time so 
as to allow a direct link with between the house of Osorkon III and 
Dyn. 25. 

A somewhat more precise knot making a temporal link between the 
Libyan and Nubian periods is possible only via Dyns. 22 and 24, and 
possible fixpoints can only be gained for Dyn. 25. 

6. The Chronological Framework for Dyn. 25 

The beginning of the reign of Taharqa lies in year 690 BC, and this 
is not disputed. 162 For a long time, his predecessor Shebitku (highest 
date is year 3 163 was assigned a reign of 8~ 1 2 years, and at the most 
13 regnal years. 164 However, the inscription of Sargon II at Tang-i Var 
reveals that Shebitku was already (at the latest) king in 706, lfo and thus 
reigned for at least 16 years. As his predecessor Shabaka ruled for at 
least 14 full years (cf. below), the beginning of his reign would be at 
the latest in 720 BC. Since one had once assumed that there were 
good reasons for believing that the Nubian rule in Egypt could not 
have begun before 716 or indeed 712 (cf. below), it was suggested a 
number of times that Shebitku was only (co)regent in Nubia while his 
senior partner, Shabaka (with dates according to his reign) ruled in 
Egypt. 166 This is historically quite improbable, aside from the fact that 
there has never been the slightest hint at any form of coregency of the 

160 Opposing D. Kessler, SAK 9 (1981), 238. 

161 On the fragment of a vessel in the Museo Barracco in Rome, cf. L. Bongrani 
Fanfoni, OrAnt 26 (1987), 65-71; pis. 2-3. 

162 Cf. TIP, §§ 130-131; Beckerath, Chronologie, 91. 

163 Nile level record no. 33, cf. most recently J.v. Beckerath, CM 136 (1993), 7-9. 

164 TIP, §§ 126; 468; Beckerath, Chronologie, 92. 

165 G. Frame, Orientalia 68 (1999), 31-57; cf. D. Kahn, Orientalia 70 (2001), 1-3. Cf. 
also N. Na'aman, N.A.B.U. 1999, Nr. 3, 63 (65). 

166 So Redford, Orientalia 68 (1999), 58-60; Beckerath, SAK 29 (2001), 3-6; Kitchen, 
in: Bietak, ed., SCIEM Haindorf 1996/98, 50-51. 


Nubian kings of Dyn. 25. Had Shabaka been ruler of Egypt in the 
year 707/706 and Shebitku his "viceroy" in Nubia, one would definitely 
expect that the opening of diplomatic relations with Assur as well as 
the capture and extradition of Yamani would have been part of Shabaka's 
responsibility. Sargon can also be expected to have named the regent 
of Egypt and senior king, rather than the distant viceroy Shebitku. If, 
on the other hand, Shebitku was already Shabaka's successor in 707/706, 
the reports of the Yamani affair become clearer and make more sense. 
It had hitherto been assumed that the Nubian king (Shabaka) handed 
over Yamani more of less immediately after his flight to Egypt. 167 Now 
it appears to be certain that Yamani was only turned over to the 
Assyrians a couple of years later. 1615 It then becomes much more prob- 
able that Shabaka awarded him asylum, but that Shebitku did not feel 
bound by his predecessor's word and that he desired to make a ges- 
ture of good will towards the Assyrians at the start of his reign, and 
that he extradited Yamani. 169 This interpretation also matches with the 
peculiar insertion into Sargon's large "ceremonial inscription" in 
Khorsabad where the king of Nubia is described as residing in a very 
distant, inaccessible land."" The formulation of his Nile level record 
(no. 33) also supports the idea that Shebitku only came to Egypt in 
his year 3. 171 

Shabaka must, therefore, have already been dead in 707/706. The 
"international" reasons which have hitherto been used to justify plac- 
ing his reign in Egypt after 716 or even 712 cannot therefore be cor- 
rect, and in fact they are wrong. The events of the years (around) 725 
(when Hosea of Israel addresses an appeal for aid to a "So, King of 
Egypt") 1 ' 2 and around 720 (when an unknown Egyptian sovereign sends 
a general named Re'e leading an army into Palestine to support a 
revolt against the Assyrians only to be defeated at Raphia), 1 ' 3 are not 

167 Cf. e.g., TIP, § 341. 

168 Cf. A. Fuchs, "Die Annalen des Jahres 711 v. Chr," State Archives of Assyria Studies 
VIII, 127-31. The actual text of the Assyrian reports recording that Yamani lived in 
Egypt "like a thief" likewise indicate a prolonged stay. 

169 Had Shabaka himself extradicted Yamani after having granted him asylum for 
years, that would have been an inconvertible sign of weakness. 

170 A. Fuchs, Die Inschriften Sargons II. aus Khorsabad (Gottingen 1994), 221-222; 348-349; 
Frame (n. 165), 53. 

171 Cf. Beckerath (n. 163), 7-9. 

172 2 Kgs. 17,4. 

173 Annals of Sargon II from Khorsabad, 11. 53—5, cf. Fuchs, Inschriften, 90; 315; cf. 
also the threshold inscriptions from Khorsabad, 11. 38-41, ibidem, 262; 360; and a 
clay cylinder from Khorsabad, 1. 19, ibidem, 34; 290. 


relevant for the dating of the Nubian rule in Egypt. 174 In the year 716, 
Sargon II extends his sphere of control further south, and receives trib- 
ute (or the like) "from Pharaoh, the king of the Land of Egypt". 173 
Another source is more precise, recording that Shilkanni, the king of 
Egypt, sent 12 large steeds as a greeting present. 176 Shilkanni could be 
Osorkon IV," 7 but he is in any case a LE and not a Nubian king. In 
the case of the Yamani-affair (711— 706) 1/8 the city of Ashdod asks 
"Pharaoh, the king of Egypt, a prince, who could not rescue it" for 
an alliance, apparentiy in vain. As the Assyrians attack, Yamani flees 
"to the border of Egypt in the area of Nubia", 179 where he lives "(secretiy) 
like a thief", until extradited by Shebitku. Neither the events of 716 
nor 7 1 1 can possibly serve as a terminus post quern for the beginning of 
Nubian rule. The pharaoh whose alliance was requested in 712/711 
can only be either Shabaka or a Delta Prince, but even in the latter 
case, it would not imply that Shabaka had not yet been recognized in 
Memphis. Shilkanni apparentiy had good reasons for trying to reassure 
the Assyrians; but this does not solve the issue of who had the upper 
hand in Egypt. In the account of his third campaign, Sennacherib 
reports that at the battle at Eltekeh (701), Hezekiah made appeals to 
"the kings of Egypt" and the troops of the King of Nubia (Shebitku). 
On this occasion, the Assyrian king captured "the charioteers and the 
sons of the kings of Egypt" and "the charioteers of the king of Nubia". 180 
From the Assyrian point of view, the enemies are perceived primarily 
as a kind of coalition, and this may have correspond to the facts, for 

174 There is one hint that Nubian soldiers took part in the battle at Raphia (cf. 
Kahn, Orientalia 70, 11-12), but these could have been mercenaries. 

175 Annals of Sargons II from Khorsabad, 11. 123—4, cf. Fuchs, Inschriften 110; 320. 

176 Assur Prisma, cf. Fuchs, Annalen, 28-29; 57. 

177 This is, however, by no means certain, a name such as Srkn or the like would 
be more reasonable, cf. J. Yoyotte, Kemi 21 (1971), 51-52. 

178 Mentioned in the following inscriptions: Annals of Sargon II from Khorsabad, 
11. 241-254, cf. Fuchs, Inschriften, 132-5; 326; small ceremonial inscription from Khorsabad, 
11. 11-14, cf. Fuchs, Inschriften, 76; 308; Orientalia 68 (1999), 52-53; large ceremonial 
inscription from Khorsabad, 11. 90-112, cf. Fuchs, Inschriften, 219-222; 348-9; Orientalia 
68, 53; Niniveh Prisma VIIB, cf. Fuchs, Annalen, 44-6; 73-4; inscription from 
Tang-i Var, cf. Frame (n. 165), 31-57. 

179 This frequently discussed phrase (cf. most recently L. Depuydt, JEA 79 [1993], 
272, n. 24; Fuchs, Inschriften, 220; 348; 452; Frame (n. 165), 52, n. 24) seems to mean 
something like "to that part of Egypt, which was under the direct control of the 

180 Cf. E. Frahm, "Einleitung in die Sanherib-Inschriften" , AfO Beiheft 26 (1997), 54; 59. 
On the 3rd campaign of Sennacheribs as a whole, cf. ibidem 10-11. 


even under Asarhaddon and Assurbanipal the princes of the Delta are 
represented as acting independently on the international stage. It is thus 
inadmissible to use evidence of such activities as a base for defining 
the beginning of Dyn. 25. 

There are no obstacles to ending the reign of Shabaka in 706 at the 
latest; on the contrary, everything suggests that Shebitku ruled alone 
from 707/706 to 690. Year 2, 3, 4, 6, 10, 12, 13, 14, 15 li!1 are doc- 
umented for Shabaka, and he is generally assigned 14 full years. An 
indirect confirmation of this can be found in Manetho, if one allows 
for a slip, 1 " 2 by assigning the 14 years Africanus gives to Shebitku to 
Shabaka. However, in view of the unreliability of the Manetho tradi- 
tion concerning Dyn. 25 183 this does not mean much. A possibility for 
calculating can also be deduced from two stelae from Kawa where 
Taharqa states that he was 20 years old when Shebitku called upon 
him to go from Nubia to Egypt. 1 " 4 As this will doubtless have taken 
place in the course of the preparations for the campaign which led to 
the battle at Eltekeh where Taharqa saw action, 185 he must have been 
born ca. 722/721. If he was a son of Piye's (as is generally assumed), 186 
the latter must have lived until at least 723 and perhaps a bit longer. 187 
However, it is by no means certain that Taharqa was really the bio- 
logical brother of Shepenupet II and thus the son of Piye. 188 Nevertheless, 
a reign of 14—15 years for Shabaka remains highly probable. Favouring 
this is also the fact that there is a relatively complete coverage of dates 
from the second decade of his reign (10, 12, 13, 14, 15), and a large 
hole would be improbable. He must thus have come to the throne at 
the latest in 720, or more probably 721 or 722. His second year would 
thus be ca. 720 (721—719), and also year 6 of Bocchoris. 189 It is calculating 

181 Block statue BM 24429, cf. Leclant, Enquetes, 15-27; pi. 5-6. 

182 Thus Beckerath, Chronologie, 92; TIP, § 421. 

183 Cf. TIP, § 468. 

184 Stela IV, 11. 7-9; V, 11. 16-7, cf. Macadam, Kawa I, 15; 28; pi. 7-10. 

185 TIP, § 127-9; 133. 

186 According to the stela of Nitokris, 11. 3-4 (cf. JEA 50 [1964], 74; pi. VIII) the 
Divine Wife Shepenupet II, a daughter of Piye, was his sister, cf. TIP, §§ 120—121. 

187 Cf. D. Kahn, Orientalia 70 (2001), 7. 

188 Cf. A. Leahy, GM 83 (1984), 43-45. 

189 An inscription from year 2 of Shabaka was found in the Serapeum, and this — 
despite some inconsistencies in the secondary literature — should be related to the same 
Apis burial as the stelae from the beginning of year 6 of Bocchoris, cf. TIP, § 114; 
J. Vercoutter, Rush 8 (1960), 62-67; PM IIP, 789. That year 2 of Shabaka was either 
the same as, or close to, year 6 of Bocchoris is clear from the sources: Manetho assigns 


the regnal years of Piye, the predecessor of Shabaka, which is uncer- 
tain, and thus likewise the link to the major campaign of year 20. 190 
In Egypt, the years 20(?), 21, 22, and 24 are documented, 191 but he is 
generally assigned a reign of 31 years as a few years must be inserted 
for Tefnakhte before the reign of Bocchoris his successor. This rests 
on the correct assumption that the various rulers of Egypt listed on 
the stela of Piye are actually identified by their rightful titles — includ- 
ing the foes of the Nubian king. If Tefnakhte is not designated a king 
there, 192 he will thus have become such only after the campaign of 
Piye. As a year 8 is recorded for Tefnakhte as king, 193 at least an addi- 
tional 7 years must have passed between Bocchoris's accession to the 
throne (ca. 725, cf. above) and the campaign of Piye, 194 and thus the 
campaign will have taken place shortly before ca. 732, perhaps 733/734. 
This is possible, but not compelling. Tefnakhte's predecessors were not 
kings, 19 ' and on two donation stelae from years 36 and 38 of Shoshenq 
V — certainly only a few years before the campaign of Piye 196 — he him- 
self does not yet bear the royal title, 197 and dates himself according to 

Bocchoris 6 years (following Africanus), and reports that Shabaka burnt him alive. 
Shabaka himself is only known in Egypt (even LE) in his regnal year 2. 

190 Only the erection of the stela with the record of this campaign is dated, in the 
first month of year 2 1 . It is generally agreed that the campaign must have taken place 
in the previous year. 

191 JEA 54 (1968), 165-172; pi. XXV; for the alleged year 30 on the mummy wrap- 
ping London BM 6640 cf. D. B. Redford, JARCE 22 (1985), 9-12; figs. 1-2, accord- 
ing to which it can be read as either 20 or 40. 

192 In 11. 19-20 he is named "Great Prince of the West", along with a few of his 
other titles; in general, however, he is merely the "Chief of the Ma" (11. 28; 80; 126). 

193 A hieratic donation stela in Athens, cf. R. el-Sayed, Documents relatifs a Sais et ses 
divinites, BdE 69 (1975), 37-53; pi. 7. K.-H. Priese {fAS 98 [1972], 19-21) and 
K. Baer (JNES 32 [1973], 23-24) have disputed that the king Tefnakhte with the 
throne-name Spss-R' is the same as the Prince Tefnakhte on the stela of Piye. They 
assume instead that this is the first king of Dyn. 26 (before the predecessor of Neco 
I) mentioned by Manetho ("Stephinates"), and thus a local prince of Sais. Opposing 
this stance is the fact that one of the stelae of Spss-R' Tefnakhte actually probably 
comes from the eastern Delta (cf. Yoyotte [n. 177], 37-40), which was most assuredly 
not under the control of the local princes ruling in Sais during Dyn. 25. Furthermore, 
Diodor (I, 43) specifies that the king Tefnakhte, predecessor of the sage Bokchoris, 
undertook an expedition to "Arabia", and this would only have been possible from 
the eastern Delta. 

194 The possibility that Tefnakhte only became king after the campaign, but that his 
regnal years were then post facto extended back to a point in time before the cam- 
paign is rejected by Kitchen (TIP, § 112). 

193 J. Yoyotte, BSFE 31 (1960), 13-22; TIP, § 113; 468. 

196 Cf. above, section 4. 

197 The unusual designation, "Great Prince of the Entire Land" reveals that his 
ambition (and certainly also his power) extended far beyond that of the ordinary Libyan 
local princes. 


Shoshenq V. If he became king shortly thereafter, e.g., after the death 
of Shoshenq V, this could only have taken place on the basis of his 
own power. As one very conscious of legitimacy, Piye would thus not 
have had the slightest reason to have designated someone as a king if 
that person had only just shortly before proclaimed himself king, and 
even less so if this person was his major opponent. 198 It would thus be 
possible to set the campaign of Piye somewhat closer to the accession 
of Bocchoris, perhaps, between 734 and 726 BC; his accession to the 
throne would thus be ca. 753-745 BC. 

7. Connecting Dyns. 22 and 25 

Shoshenq V died before the campaign of Piye, but not long before, 
since Tefnakhte claims the title "Great Prince of the Entire Land" in 
year 38. On the other hand, however, Shoshenq's rule was apparently 
uncontested in Memphis in his year 37, and thus Tefnakhte's expan- 
sion was not as advanced as at the beginning of the campaign of Piye. 
In addition, there may be another king Petubaste (cf. above) to insert 
before Osorkon IV who reigned in Bubastis and Tanis during the cam- 
paign. A period of about 5 years between the death of Shoshenq V 
and year 20 of Piye would appear reasonable. 

For the kings from Shoshenq I to Takelot I we can reckon at least 
21 + 35+15 years, for Osorkon II at least 30 years, as a year 29 is 
very probably documented and the genealogical data favours a long 
reign (cf. above, section 1). For the kings Shoshenq III, Shoshenq Ilia, 
Pami and Shoshenq V, we have made a minimal period of 95 years 
(cf. above, section 3). If we start with year 945 as the beginning of 
Dyn. 22, the year 38 of Shoshenq V is to be set in 749 at the earliest. 
The campaign of Piye can be placed in the years 734—726 (cf. above, 
section 6); 10-18 years would remain to bridge the period between the 
campaign and year 38 (+ 5) of Shoshenq V. This result is also realis- 
tic as those reigns the duration of which is not certain were assigned 
minimal values here. Where these missing years must be placed is a 

198 On his "small stela" (Khartoum 1851, G. A. Reisner, %AS 66 [1931) 89-100; 
pis. V— VI) Piye clearly enunciates that only that person is king whom he makes king, 
and not those whom he forbids. This could apply to Tefnakhte, nor does he call 
Tefnakhte "Great Prince of the Entire Land", but rather "Chief of the Ma". 


matter of speculation, but there are several possibilities. 199 First of all, 
the year 945 is not certain. If the campaign of Shoshenq in Palestine 
(926/925 in year 5 of Rehabeam) did not take place in his year 20, 
but rather a few years earlier — as is entirely possible 200 — the beginning 
of the reign must accordingly be placed somewhat later. Candidates 
for a somewhat longer reign are Shoshenq V himself and Osorkon II. 
Likewise, Pami may have reigned for more than 6 years, 201 if his "annals" 
were not written posthumously (cf. above, section 3), and Osorkon I 
and his successors could have ruled longer than we have assumed 
above. 202 In any case, a very slight extension of a few reigns is just as 
unproblematic as setting the beginning of Dyn. 22 marginally later in 

8. Conclusion 

For the chronology of the TIP, Egyptian sources only supply the year 
690 as a certain point of departure. Additionally, the date of the cam- 
paign of Shoshenq I, presumably towards the end of his reign, can be 
placed with the aid of Near Eastern chronology in 925/926. 2()3 Between 
these two there is not one single firm date, but the sequence of kings 
and the highest known dates for these kings does not leave significant 
gaps. The general framework of the chronology of this age is certain. 
Additional finds of dated monuments from this period will hopefully 
add to the previous discoveries, and lead to an even higher degree of 
resolution, leaving still less uncertainty. 

199 Assigning the entire sum of years to the reign of Osorkon II, as Aston (n. 31; 
145-148) does, is not necessarily the most logical possibility. 

200 Cf. above, Jansen-Winkeln, Chapter II. 9. 

201 Beckerath, Chronologie, 98, assigns him 1 1 years. 

202 The usual numbers still depend to a great extent upon the very doubtful figures 
for this period provided by the copyists of Manetho. 

203 Cf. above, n. 200. Following alternative and acceptable calculations in OT stud- 
ies, the year 5 of Rehabeam would not have been 926/5, but rather 922/1 (H. Donner, 
Geschichte des Volkes Israel und seiner Nachbam in Grundziigen, 2 [1995], 274); and this would 
correspond to the Egyptian dates quite well. A "chronological problem" noted by 
Donner ibidem, 321, n. 14) does not exist in this fashion: the Egyptian chronology is 
absolutely dependent upon Near Eastern chronology. If one follows Begrich/Jepsen 
and not Thiele, one simply shifts the accession of Shoshenq I by the same margin. 




Leo Depuydt 

In the period at hand, events can dated exactly in absolute terms, that 
is, in relation to the present moment in time. Thus, the death of 
Psammetichus II in Month 1 Day 23 of his Year 7, that is, on 9 Feb 
589 BC, is separated from the same time of day (whatever it was) on 

I Jan AD 2003 by 9463 1 1 full days or 24-hour periods. Chapter III. 

I I outlines the general principles of day-exact chronology, with refer- 
ences to contributions that provide more detail. 1 In terms of chrono- 
logical structure, the period has three natural subdivisions, each with 
its own anatomy: (1) Dyn. 26, the Saite dynasty; (2) Dyn. 27, consist- 
ing of Persian rulers; and (3) the fourth century BC up to Alexander's 
conquest of 332 BC, when Egypt was independent yet in Persia's sphere 
of influence. Accordingly, the following three equations have been 
obtained in fundamentally different ways. 

1. Year 7 Month 1 Day 23 (I fyt 23) = 9 Feb 589 BC 
of Psammetichus II 

2. Year 15 Month 4 Day 16 (16 Hathyr) = 4 Mar 471 BC 
of Xerxes I 

3. Year 16 Month 8 Day 21 (21 Pharmouthi) = 5 Jul 343 BC 
of Nectanebo II 

The structural hierarchy of the chronology of Dyns. 26—31 is as fol- 
lows. Dyn. 27 is most secure, ultimately owing to synchronies with 
Mesopotamia established through Babylonian astronomical texts. As the 
anchor, Dyn. 27 is treated here first. Dyn. 26 depends on Dyn. 27 and 
is reasonably secure. It is treated next, along with Taharqa's reign, the 
last of Dyn. 25, which depends on Dyn. 26. The fourth century BC 
is best seen as a gap to be bridged between Dyn. 27 and Alexander's 
conquest of 332 BC. It is only partly day-exact. 

1 For a brief outline, see Beckerath, Chronologk, 79-88, with supplements in L. Depuydt, 
WdO 30 (1999), 143-51, a review. See also H.-J. Thissen, "Chronologie der fruhde- 
motischen Papyri", Enchoria 10 (1980), 105-25. 


1. Dynasty 27 

The four pioneers who in the more recent past have done the most 
to cement the chronology of Egypt in 664 BC— 332 BC by consolidat- 
ing Dyn. 27 as its anchor are E. Meyer, R. A. Parker, A. Sachs, and 
above all F. X. Kugler. 2 By manipulating (1) Ptolemy's Canon, (2) pre- 
dating of postdating, and (3) the cuneiform record, 3 the following approx- 
imate dates for the beginnings of the reigns are obtained. Details follow 
in the notes to the table in section 4. Day dates derived from ancient 
lunar dates can be one to two days off. 

Cambyses Aug 530 BC 

Darius I 29 Sep-22 Dec 522 BC 

Xerxes I late Nov 486 BC 

Artaxerxes I 5 Aug 465 BC-2 Jan 464 BC 

Darius II 25 Dec 424 BC-13 Feb 423 BC 

Artaxerxes II 18 Sept 405 BC-9 Apr 404 BC 

Babylonian regnal Year 1 began on the first Babylonian new year fol- 
lowing these dates (see 1.2 in Chapter III. 11), that is, around the 
spring equinoxes of 529, 521, 485, 464, 423, and 404 BC. The num- 
berless period lasting from the beginning of the reign to the first Baby- 
lonian new year may be called the accession year. The above dates 
can be translated as follows into reign lengths detailing years only. 

Dyn. 27 527-525 ca. 400 

Cambyses (second half of reign) 527-525 522 

Darius I 522 486 

Xerxes I 486 465 

Artaxerxes I 465 424/3 

Darius II 424/3 405/4 

Artaxerxes II (beginning of reign) 405/4 ca. 400 

En-dashes (— ) denote time-periods. Slashes (/) mark ancient calendar-years straddling 
two Julian years. 

2 Pioneering contributions are: E. Meyer, Forschungen zur Altai Geschichte, II: Z ur 
Geschichte des funften Jahrhunderts v. Chr. (Halle, 1899); F. X. Kugler, und Stemdienst 
in Babel (Miinster, 1907-24); R. A. Parker, "Persian and Egyptian Chronology", AJSL 
58 (1941), 285-301; id., "Darius and His Egyptian Campaign", AJSL 58 (1941), 373-377; 
Parker & Dubberstein, Chronology; A. Sachs's work on the Babylonian diaries. 

3 See sections 2.1, 1.2, and 2.3 in Chapter III. 11. 


2. Dynasty 26 

The main structural feature of the day-exact chronology of Dyn. 26 is 
its dependence on the day-exact chronology of Dyn. 27. Already in the 
nineteenth century AD, data from Greek historians and Serapeum 
stelae (see 2.2 in Chapter III. 1 1) had fixed the reigns of Dyn. 26 
absolutely to within one or two years. In the 1950s, two contributions 
by R. A. Parker pushed the limit of day-exact chronology back from 
525 BC in two moves: (1) from 525 BC back to 664 BC; (2) from 664 
BC back to 690 BC. Each move rests on a single piece of evidence. 4 

(1) A civil-lunar double date deciphered jointly by M. Malinine and 
R. A. Parker at Brown University in a photograph of the abnormal 
hieratic papyrus Louvre 7848 equates civil Month 10 Day 13 (II Shemu 
13) with lunar Month 9 Day 15 (I Shemu 15) in Year 12 of Amasis. 
Lunar Day 15 ought to fall around full moon. Before this date sur- 
faced, Amasis's Year 12 had mostly been equated with the 365-day 
year 10 Jan 558 BC-9 Jan 557 BC, in which civil Month 10 Day 13 
equals 19 October. However, 19 Oct 558 BC is not close to full moon. 
In the preceding Egyptian year, 10 Jan 559 BC-9 Jan 558 BC, civil 
Month 10 Day 13 also equals 19 Oct. and 19 Oct 559 BC does occur 
near the full moon of 21 Oct 12:09PM.' The earlier Egyptian year is 
therefore in all probability Amasis's Year 12. 6 

This backward shift of Amasis's regnal years affects the dating of 
Cambyses's conquest. Traditionally, the conquest had been dated to 
spring 525 BC and Amasis's Year 44 had been considered his last. But 
now, Day 1 of Year 44 was moved back from 2 Jan 526 BC to 2 Jan 
527 BC. To keep the end of Amasis's reign close to the conquest, 
R. A. Parker postulated an unattested Year 45 for Amasis beginning 
on 2 Jan 526 BC. However, postulating a Year 45 was rendered unnec- 
essary when the arguments dating the conquest to spring 525 BC. 

4 A passage in Demotic papyrus Berlin 13588 that has been interpreted variously 
as a solar eclipse (E. Hornung, "Die Sonnenfinsternis nach dem Tode Psammetichs 
I", ZAS 92 (1966), 38-9) and as a lunar eclipse (M. Smith, "Did Psammetichus I Die 
Abroad?", OLP 22 (1991), 101-9) has possible chronological relevance (see also L. Depuydt, 
"On the Consistency of the Wandering Year as Backbone of Egyptian Chronology", 
JARCE 32 (1995), 43-58, at 53, note 50); for a different view, see Chapter III. 4. 

' Goldstine's time, Moons, 37, for Babylon, minus 47 minutes for Thebes. 

6 R. A. Parker, "The Length of Reign of Amasis and the Beginning of the Twenty- 
sixth Dynasty", MDAIK 15 (1957), 208-12. pLouvre 7848 has now been published by 
K. Donker van Heel, Abnormal Hieratic and Early Demotic Texts Collected by the Theban 
Choachytes in the Reign of Amasis (Leiden, 1995), 93-99. 


became invalid. The conquest is now better dated more imprecisely to 
527-525 BC. 7 

(2) Serapeum stela Louvre IM 3733, the official epitaph of an Apis 
bull born in Taharqa's Year 26 and deceased in Psammetichus I's Year 
20 pushes day-exact chronology back further from 664 BC to 690 BC. K 
Accordingly, Taharqa would have counted his reign from some day in 
the Egyptian year 12 Feb 690 BCM 1 Feb 689 BC. The year 12 Feb 
691 BC-11 Feb 690 BC is also a possibility. This window of doubt 
exists because IM 3733 lacks certain information: (1) the months and 
days of the bull's life-span, given as 2 1 years; (2) the month and day 
of his birth in Taharqa's Year 26; and (3) the year of his day of instal- 
lation, which was Month 8 Day 9. 9 The lengths of the reigns of Dyn. 
26 are as follows. Details follow in the notes to the table in section 4. 

Dyn. 26 664/3 ca. 527-5 

Psammetichus I 664/3 610 

Necho II 610 595 

Psammetichus II 595 589 

Apries 589 570 

Amasis 570 527/6 

Psammetichus III? some (6?) months in 527-525? 

En-dashes (— ) denote time-periods. Slashes (/) mark ancient calendar-years straddling 
two Julian years. 

3. Dynasties 28 to 31 

The main structural feature of the chronology of Egypt in the fourth 
century BC up to Alexander's conquest of 332 BC is that it is partly 
day-exact. This feature deserves attention first. The evidence is in 
Nectanebo's Dream, which prophesies the demise of Egypt's last native 
ruler. The king of the Greek version had been identified with Nectanebo II, 

' For a detailed discussion, see L. Depuydt, "Egyptian Regnal Dating under Cambyses 
and the Date of the Persian Conquest", in: Studies in Honor of William Kelly Simpson 
(Boston, 1996), 179-190, at 184, note 23. See now also J.v. Beckerath, "Nochmals die 
Eroberung Agyptens durch Kambyses", %AS 129 (2002), 1-5; and E. Cruz-Uribe, "The 
Invasion of Egypt by Cambyses", Transeuphratene 25 (2003), 9-60, at 54-57. 

8 R. A. Parker, "The Length of Reign of Taharqa", Rush 8 (1960), 267-269. 

9 For details, see Depuydt, JARCE 32, 52-53. 


not Nectanebo I, but it took a Demotic fragment to confirm this 
definitively. 10 In the Greek version in papyrus Leiden I 396, the Dream 
is dated to the night from 21 to 22 Pharmouthi (Day 21 to Day 22 
of Month 8) of Nectanebo IPs Year 16." It was about full moon (tcaia 
Geov 8ia 8exo|ieviav). The lunar date can be identified with the help 
of Manetho's king-lists. In Manetho, 12 about 72 to 74 years separate 
the end of the reign of Darius II in 405/4 BC from Alexander's con- 
quest of 332 BC. Nectanebo IPs Year 16 is therefore about 62 years 
away from 405/4 BC, or falls about halfway between 345 BC and 340 
BC. By far the closest match for the combination of full moon with 
21/2 Pharmouthi falls in 343 BC. In this year, 21/2 Pharmouthi equals 
5/6 Jul 343 BC. Full moon occurred about noon of 6 July. 13 The fol- 
lowing equations result. Details follow in the table in section 4. 

Nectanebo II Year 1 21 Nov 359-20 Nov 358 

Nectanebo II Year 2 21 Nov 358-19 Nov 357 
And so on 

Nectanebo II Year 19 16 Nov 341-15 Nov 340 

Nectanebo II Year 20 16 Nov 340-15 Nov 339 

It is not known when in the Egyptian year 21 Nov 359 BC— 20 Nov 
358 BC Nectanebo II assumed power. According to Manetho, he reigned 
20 years. The highest regnal date in hieroglyphic sources is Year 18. 14 
According to Manetho, or an addition to Manetho, Artaxerxes III, 
also called Ochos, conquered Egypt in his Year 20. '' His Babylonian 
Year 20 ran from spring 339 BC to spring 338 BC. According to 
predating of postdating (see 1.2 in Chapter III. 11), his Egyptian 
year 20 would begin on the Egyptian new year of 16 Nov 340 BC. 

10 K. Ryholt, "A Demotic Version of Nectanebos' Dream (P. Carlsberg 562)", J(TE 
122 (1998), 197-200, with bibliography. Cf. also A. Spalinger, "The Date of the Dream 
of Nectanebo", SAK 19 (1992), 295-304. 

11 The Demotic version's Year 18 must be an error. Indeed, three other Demotic 
versions, Carlsberg 424, 499, and 559, date a sequel to the Dream to Year 16 (Ryholt, 
ZPE 122, 198). 

12 Waddell, Manetho. 

13 Goldstine's time [Moons, 55) for Babylon, minus 53 minutes for Memphis. 

14 Kienitz, Geschichte, 216 (part of a useful and extensive listing of documents dating 
to the fourth century BC). 

15 Waddell, Manetho, 184-187. 


Remarkably, the conquest is now mostly dated earlier, to 343/2 BC. 16 
But on closer inspection, the sole argument ever adduced in favor of 
343/2 BC appears non-binding. Persian envoys visited Athens when 
Lukiskos was archon, from July 344 BC to July 343 BC. It has been 
assumed that the army of Greek mercenaries hired on the occasion 
could not have been maintained long without action. The invasion of 
Egypt must therefore have followed soon, in the fall of 343 BC. 17 This 
argument is reasonable, but hardly conclusive. It contradicts other evi- 
dence that seems firmer. 340/39 BC remains preferable as the date of 
the re-conquest. 

For the rest of Dyns. 28 to 30, Manetho serves as the basis, fine- 
tuned by data from the monuments. Current assignments of regnal 
years to Egyptian wandering years are not definitive, but in all prob- 
ability correct to within one or two years. Africanus, generally consid- 
ered the best source for Manetho, who is himself not preserved, gives 
73 years and 4 months for that period: 6 years for Dyn. 28; 20 years 
4 months for Dyn. 29; 38 years for Dyn. 30; and 9 years for Dyn. 31. 
Provisional dates for the reigns are as follows. 18 

Dyn. 28 

Amyrtaios about 404/3-398/7 

Dyn. 29 

Nepherites I about 398/7-392/1 

Achoris about 392/1-379/8 

Psammuthis brief reign 

Nepherites II brief reign 

Dyn. 30 

Nectanebo I (Nectanebes) about 379/8-361/0 

Teos/Tachos about 361/0-359/8 

Nectanebo II (Nectanebos) about 359/8-342/1 

16 Kienitz, Geschkhk, 172; A. B. Lloyd, "Egypt, 404-332 BC", in: CAH 2 , vol. 6 
(Cambridge, 1994), 337-360, 981-987, at 359, where the date is deemed a 'fact'. 

17 E. J. Bickerman, "Notes sur la chronologie de la XXX r dynastie", in: Melanges 
Maspero I: Orient Ancien (Cairo: MIFAO 56, 1934), 77-84, at 81. 

18 From Lloyd, CAH 2 , vol. 6, 358. Lloyd's survey of Egypt in 404-332 BC includes 
an appendix on chronology. See also J. D. Ray, "Psammuthis and Hakoris", JEA 72 
(1986), 149-58; C. Traunecker, "Essai sur l'histoire de la XXIX C Dynastie", 79 (1979), 
395-436. M. Chauveau ("Les archives d'un temple des oasis au temps des Perses", 
BSFE 137 (1996), 32-47) suggests that Amyrtaios, sole ruler of Dyn. 28, sometimes 
used the name Psammetichus. That would him make the fifth ruler of that name, in 
addition to Psammetichus I and II of Dyn. 26 and two postulated ephemeral rulers 
dating to about 526 BC and to about the mid 480s BC. 


The evidence for Dyn. 31 is sparse. 19 The following dates for the lengths 
of the Egyptian reigns of its three Persian kings are derived from the 
date of Artaxerxes Ill's conquest (see section 3 above) and from Ptolemy's 
Canon (see 2.1 in Chapter III. 11). The first Babylonian regnal years 
of Arses and Darius III began in the spring of 337 BC and of 335 BC 
respectively, that is, on the first Babylonian new year following the 
Canon's beginnings for their reigns, namely 16 Nov 338 BC and 15 
Nov 336 BC. Their actual reigns may have begun up to a year earlier 
and therefore have either preceded or followed the Canon's beginning. 


4. Regnal Years in 690 BC-332 BC 

All the years in the following table are exactly 365 days long. The 
years printed in italics include a Julian 29 February. The Era of 
Nabonassar in the first column is a year-count from the first king of 
Ptolemy's Canon (see 2.1 in Chapter III. 11). The table joins the Canon 
in 525 BC. Details on the transitions of the historical reigns appear in 
the notes. The years from 525 BC to the end of the fifth century BC 
are almost certainly also the actual historical Egyptian regnal years; the 
years back to 664 BC, with high probability; those back to 690 BC, 
quite possibly. For the fourth century BC, only regnal years placed 
with reasonable certainty are listed. To convert, say, Month 5 Day 29 
of Year 5 of Cambyses, one may proceed as follows. 

Dyn. 31 

ca. 340 

Artaxerxes III (end of reign) 

ca. 340 



Darius III 


19 See D. Devauchelle, "Reflexions sur les documents egyptiens dates de la Deuxieme 
Domination perse", Transeuphratene 10 (1995), 35-43; A. B. Lloyd, "Manetho and the 
Thirty-first Dynasty", in: Pyramid Studies and Other Essays Presented to I. E. S. Edwards 
(London: EES Occasional Publications 7, 1988), 154—160; A. Spalinger, "The Reign 
of King Chababash: An Interpretation", £AS 105 (1978), 142-54. On the Demotic 
Chronicle, see J. H. Johnson, "The Demotic Chronicle as an Historical Source", Enchoria 
4 (1974), 1-17. 




2 Jan 525 


Month 1 Day 1 

+ 29 

+ 29 

31 Jan 525 


Month 1 Day 3 

+ 1 

+ 1 

1 Feb 525 


Month 2 Day 1 

+ 28 

+ 28 

29 Feb 525 


Month 2 Day 29 

+ 1 

and so on 

+ 1 

Era of 


Length of 365-day 


Regnal Year 

Egyptian Year 


Shabataka? Year ? 

+ Taharqa 


12 Feb 690-11 Feb 689' 




12 Feb 689-10 Feb 688 




11 Feb 688-10 Feb 687 




11 Feb 687-10 Feb 686 




11 Feb 686-10 Feb 685 




11 Feb 685-9 Feb 684 




10 Feb 684-9 Feb 683 




10 Feb 683-9 Feb 682 




10 Feb 682-9 Feb 681 




10 Feb 681-8 Feb 680 




9 Feb 680-8 Feb 679 




9 Feb 679-8 Feb 678 




9 Feb 678-8 Feb 677 




9 Feb 677-7 Feb 676 




8 Feb 676-7 Feb 675 




8 Feb 675-7 Feb 674 




8 Feb 674-7 Feb 673 




8 Feb 673-6 Feb 672 




1 Feb 672-6 Feb 671 




7 Feb 671-6 Feb 670 




7 Feb 670-6 Feb 669 




7 Feb 669-5 Feb 668 




6 Feb 668-5 Feb 667 




6 Feb 667-5 Feb 666 




6 Feb 666-5 Feb 665 




6 Feb 665-4 Feb 664 




5 Feb 664- 

+ Psammetichus I 1 

-4 Feb 663" 


Psammetichus I 2 

5 Feb 663-4 Feb 662 


Psammetichus I 3 

5 Feb 662-4 Feb 661 


Psammetichus I 4 

5 Feb 661-3 Feb 660 

{continued on next page) 



Table (cont.) 

Era of 


Length of 365-day 


Regnal Year 

Egyptian Year 



[ 5 

4 Feb 

660-3 Feb 659 



[ 6 

4 Feb 

659-3 Feb 658 



[ 7 

4 Feb 

658-3 Feb 657 




4 Feb 657-2 Feb 656 




3 Feb 

656-2 Feb 655 




3 Feb 655-2 Feb 654 



L 11 

3 Feb 654-2 Feb 653 



[ 12 

3 Feb 653-1 Feb 652 



[ 13 

2 Feb 652-1 Feb 651 



[ 14 

2 Feb 651-1 Feb 650 



[ 15 

2 Feb 650-1 Feb 649 



[ 16 

2 Feb 649-31 Jan 648 



[ 17 

1 Feb 648-31 Jan 647 



[ 18 

1 Feb 647-31 Jan 646 



[ 19 

1 Feb 646-31 Jan 645 



[ 20 

1 Feb 645-30 Jan 644 



[ 21 

31 Jan 

644-30 Jan 643 



[ 22 

31 Jan 

643-30 Jan 642 



[ 23 

31 Jan 

642-30 Jan 641 



[ 24 

31 Jan 

641-29 Jan 640 



[ 25 

30 Jan 

640-29 Jan 639 



[ 26 

30 Jan 

639-29 Jan 638 



[ 27 

30 Jan 

638-29 Jan 637 



[ 28 

30 Jan 

637-28 Jan 636 



[ 29 

29 Jan 

636-28 Jan 635 



[ 30 

29 Jan 

635-28 Jan 634 



[ 31 

29 Jan 

634-28 Jan 633 



[ 32 

29 Jan 

633-27 Jan 632 



[ 33 

28 Jan 

632-27 Jan 631 



[ 34 

28 Jan 

631-27 Jan 630 



[ 35 

28 Jan 

630-27 Jan 629 



[ 36 

28 Jan 

629-26 Jan 628 



[ 37 

27 Jan 

628-26 Jan 627 



[ 38 

27 Jan 

627-26 Jan 626 



[ 39 

27 Jan 

626-26 Jan 625 



[ 40 

27 Jan 

625-25 Jan 624 



[ 41 

26 Jan 

624-25 Jan 623 



[ 42 

26 Jan 

623-25 Jan 622 



[ 43 

26 Jan 

622-25 Jan 621 



[ 44 

26 Jan 

621-24 Jan 620 



[ 45 

25 Jan 

620-24 Jan 619 




25 Jan 

619-24 Jan 618 

{continued on next page) 



Table (cont.) 

Era of 

Regnal Year 

Length of 365-day 
Egyptian Year 







Psammetichus I 47 
Psammetichus I 48 
Psammetichus I 49 
Psammetichus I 50 
Psammetichus I 51 
Psammetichus I 52 
Psammetichus I 53 
Psammetichus I 54 
Psammetichus I 55 

Necho II 1 

Necho II 2 

Necho II 3 

Necho II 4 

Necho II 5 

Necho II 6 

Necho II 7 

Necho II 8 

Necho II 9 

Necho II 10 

Necho II 11 

Necho II 12 

Necho II 13 

Necho II 14 

Necho II 15 

Necho II 16 

+ Psammetichus II 1 

Psammetichus II 2 
Psammetichus II 3 
Psammetichus II 4 
Psammetichus II 5 
Psammetichus II 6 
Psammetichus II 7 
+ Apries 1 
Apries 2 
Apries 3 
Apries 4 
Apries 5 
Apries 6 

25 Jan 618-24 Jan 617 
25 Jan 617-23 Jan 616 
24 Jan 616-23 Jan 615 
24 Jan 615-23 Jan 614 
24 Jan 614-23 Jan 613 
24 Jan 613-22 Jan 612 
23 Jan 612-22 Jan 611 
23 Jan 611-22 Jan 610 
23 Jan 610- 

-no later than 18 Nov 6 1 iis 
by 19 Nov 610 at the latest iiL - 

-22 Jan 


23 Jan 609-21 Jan 

22 Jan 608-21 Jan 

22 Jan 607-21 Jan 

22 Jan 606-21 Jan 

22 Jan 605-20 Jan 

21 Jan 604-20 Jan 

21 Jan 603-20 Jan 602 

21 Jan 602-20 Jan 601 

21 Jan 601-19 Jan 600 

20 Jan 600-19 Jan 599 

20 Jan 599-19 Jan 598 

20 Jan 598-19 Jan 597 

20 Jan 597-18 Jan 596 

19 Jan 596-18 Jan 595 

19 Jan 595- 

-at least until 4 May 595" 

by 23 Nov 595 at the latest- 

-18 Jan 594 iv 
19 Jan 594-18 Jan 593 
19 Jan 593-17 Jan 592 
18 Jan 592-17 Jan 591 
18 Jan 591-17 Jan 590 
18 Jan 590-17 Jan 589 
18 Jan 589-9 Feb 589 
10 Feb 589-16 Jan 588 v 
17 Jan 588-16 Jan 587 
17 Jan 587-16 Jan 586 
17 Jan 586-16 Jan 585 
17 Jan 585-15 Jan 584 
16 Jan 584-15 Jan 583 

(continued on next page) 



Table (cont.) 

Era of 


Regnal Year 

Length of 365-day 
Egyptian Year 



i 7 


Apries 8 


Apries 9 


Apries 10 


Apries 1 1 


Apries 12 


Apries 13 


Apries 14 


Apries 15 


Apries 16 


Apries 17 


Apries 18 


Apries 19 


Apries 20 

+ Amasis 1 



s 2 



s 3 



s 4 



s 5 



s 6 



s 7 



s 8 



s 9 



s 10 



s 11 



s 12 



s 13 



s 14 



s 15 



s 16 



s 17 



s 18 



s 19 



s 20 



s 21 



s 22 



s 23 



s 24 



s 25 



is 26 

16 Jan 583-15 Jan 582 

16 Jan 582-15 Jan 581 

16 Jan 581-14 Jan 580 

15 Jan 580-14 Jan 579 

15 Jan 579-14 Jan 578 

15 Jan 578-14 Jan 577 

15 Jan 577-13 Jan 576 

14 Jan 576-13 Jan 575 

14 Jan 575-13 Jan 574 

14 Jan 574-13 Jan 573 

14 Jan 573-12 Jan 572 

13 Jan 572-12 Jan 571 

13 Jan 571-12 Jan 570 

13 Jan 570- 

-at least until 19 Oct 570 vl 

by Day x in 1 1 Jul-9 Aug 570 

at the latest vl -12 Jan 569 

13 Jan 569-11 Jan 568 

12 Jan 568-11 Jan 567 

12 Jan 567-11 Jan 566 

12 Jan 566-11 Jan 565 

12 Jan 565-10 Jan 564 

11 Jan 564-10 Jan 563 

11 Jan 563-10 Jan 562 

11 Jan 562-10 Jan 561 

11 Jan 561-9 Jan 560 

10 Jan 560-9 Jan 559 

10 Jan 559-9 Jan 558 

10 Jan 558-9 Jan 557 

10 Jan 557-8 Jan 556 

9 Jan 556-8 Jan 555 

9 Jan 555-8 Jan 554 

9 Jan 554-8 Jan 553 

9 Jan 553-7 Jan 552 

8 Jan 552-7 Jan 551 

8 Jan 551-7 Jan 550 

8 Jan 550-7 Jan 549 

8 Jan 549-6 Jan 548 

1 Jan 548-6 Jan 547 

7 Jan 547-6 Jan 546 

7 Jan 546-6 Jan 545 

7 Jan 545-5 Jan 544 

{continued on next page) 



Table (cont.) 

Era of 


Length of 365-day 


Regnal Year 

Egyptian Year 



s 27 

6 Jan 544-5 Jan 543 



s 28 

6 Jan 543-5 Jan 542 



s 29 

6 Jan 542-5 Jan 541 



s 30 

6 Jan 541-4 Jan 540 



s 31 

5 Jan 540-4 Jan 539 



s 32 

5 Jan 539-4 Jan 538 



s 33 

5 Jan 538-4 Jan 537 



s 34 

5 Jan 537-3 Jan 536 



s 35 

4 Jan 536-3 Jan 535 



s 36 

4 Jan 535-3 Jan 534 



s 37 

4 Jan 534-3 Jan 533 



s 38 

4 Jan 533-2 Jan 532 



s 39 

3 Jan 532-2 Jan 531 



s 40 

3 Jan 531-2 Jan 530 



s 41 

3 Jan 530-2 Jan 529 



s 42 

3 Jan 529-1 Jan 528 



s 43 

2 Jan 528-1 Jan 527 



s 44 

2 Jan 527- 

+ uncertain™ 

-1 Jan 526 



2 Jan 526-1 Jan 525 



2 Jan 525- 

+ Cambyses 5 

-31 Dec 525 


Cambyses 6 

1 Jan 524-31 Dec 524 


Cambyses 7 

1 Jan 523-31 Dec 523 

226 ix 

Cambyses 8 

1 Jan 522-death in 522 x 

+ Darius I's 

accession in 522 xl — 

accession year 

-31 Dec 522 

22? x 

Darius I's 

1 Jan 521- 

accession year or Year 1? 

-ca. 13 Apr 521™ 

Darius I 1 

ca. 13 Apr 521- 

-30 Dec 521 


Darius I 2 

31 Dec 521-30 Dec 520 


Darius I 3 

31 Dec 520-30 Dec 519 


Darius I 4 

31 Dec 519-30 Dec 518 


Darius I 5 

31 Dec 518-29 Dec 517 


Darius I 6 

30 Dec 517-29 Dec 516 


Darius I 7 

30 Dec 516-29 Dec 515 


Darius I 8 

30 Dec 515-29 Dec 514 


Darius I 9 

30 Dec 514-28 Dec 513 


Darius I 10 

29 Dec 513-28 Dec 512 


Darius 111 

29 Dec 512-28 Dec 511 


Darius I 12 

29 Dec 511-28 Dec 510 

(continued on next page) 



Table (cont.) 

Era of 


Length of 365-day 


Regnal Year 

Egyptian Year 



us I 13 

29 Dec 510-27 Dec 509 



us I 14 

28 Dec 509-27 Dec 508 



us I 15 

28 Dec 508-27 Dec 507 



us I 16 

28 Dec 507-27 Dec 506 



us I 17 

28 Dec 506-26 Dec 505 



us I 18 

27 Dec 505-26 Dec 504 



us I 19 

27 Dec 504-26 Dec 503 



us I 20 

27 Dec 503-26 Dec 502 



us I 21 

27 Dec 502-25 Dec 501 



us I 22 

26 Dec 501-25 Dec 500 



us I 23 

26 Dec 500-25 Dec 499 



us I 24 

26 Dec 499-25 Dec 498 



us I 25 

26 Dec 498-24 Dec 497 



us I 26 

25 Dec 497-24 Dec 496 



us I 27 

25 Dec 496-24 Dec 495 



us I 28 

25 Dec 495-24 Dec 494 



us I 29 

25 Dec 494-23 Dec 493 



us I 30 

24 Dec 493-23 Dec 492 



us I 31 

24 Dec 492-23 Dec 491 



us I 32 

24 Dec 491-23 Dec 490 



us I 33 

24 Dec 490-22 Dec 489 



us I 34 

23 Dec 489-22 Dec 488 



us I 35 

23 Dec 488-22 Dec 487 


Darius I 36 

23 Dec 487- 

-at least until 24 Nov 486" v 

+ Xerxes I's 

at the latest by 1 Dec 486- 

accession year 

-22 Dec 486 


Xerxes I's 

23 Dec 486- 

accession year or Year 1? 

-ca. 5 Apr 485" 

Xerxes I 1 

ca. 5 Apr 485"'-21 Dec 485 


Xerxes I 2 

22 Dec 485-21 Dec 484 


Xerxes I 3 

22 Dec 484-21 Dec 483 


Xerxes I 4 

22 Dec 483-21 Dec 482 


Xerxes I 5 

22 Dec 482-20 Dec 481 


Xerxes I 6 

21 Dec 481-20 Dec 480 


Xerxes I 7 

21 Dec 480-20 Dec 479 


Xerxes I 8 

21 Dec 479-20 Dec 478 


Xerxes I 9 

21 Dec 478-19 Dec 477 


Xerxes I 10 

20 Dec 477-19 Dec 476 


Xerxes 111 

20 Dec 476-19 Dec 475 


Xerxes I 12 

20 Dec 475-19 Dec 474 


Xerxes I 13 

20 Dec 474-18 Dec 473 

{continued on next page) 



Table (cont.) 

Era of 


Length of 365-day 


Regnal Year 

Egyptian Year 


Xerxes I 14 

19 Dec 473-18 Dec 472 


Xerxes I 15 

19 Dec 472-18 Dec 471 


Xerxes I 16 

19 Dec 471-18 Dec 470 


Xerxes I 17 

19 Dec 470-17 Dec 469 


Xerxes I 18 

18 Dec 469-17 Dec 468 


Xerxes I 19 

18 Dec 468-17 Dec 467 


Xerxes I 20 

18 Dec 467-17 Dec 466 

283 m 

Xerxes I 21 

18 Dec 466- 

-Day x in 4-8 Aug 465 x "' 

+ Artaxerxes I's 

by 5 Aug 465 at the earliesr™ 11 - 

accession year? 

-16 Dec 465 


Artaxerxes I's 

17 Dec 465- 

accession year or Year 1? 

-ca. 13 Apr 464 xix 

Artaxerxes I 1 

ca. 13 Apr 464-16 Dec 464 


Artaxerxes I 2 

17 Dec 464-16 Dec 463 


Artaxerxes I 3 

17 Dec 463-16 Dec 462 


Artaxerxes I 4 

17 Dec 462-15 Dec 461 


Artaxerxes I 5 

16 Dec 461-15 Dec 460 


Artaxerxes I 6 

16 Dec 460-15 Dec 459 


Artaxerxes I 7 

16 Dec 459-15 Dec 458 


Artaxerxes I 8 

16 Dec 458-14 Dec 457 


Artaxerxes I 9 

15 Dec 457-14 Dec 456 


Artaxerxes I 10 

15 Dec 456-14 Dec 455 


Artaxerxes 111 

15 Dec 455-14 Dec 454 


Artaxerxes I 12 

15 Dec 454-13 Dec 453 


Artaxerxes I 13 

14 Dec 453-13 Dec 452 


Artaxerxes I 14 

14 Dec 452-13 Dec 451 


Artaxerxes I 15 

14 Dec 451-13 Dec 450 


Artaxerxes I 16 

14 Dec 450-12 Dec 449 


Artaxerxes I 17 

13 Dec 449-12 Dec 448 


Artaxerxes I 18 

13 Dec 448-12 Dec 447 


Artaxerxes I 19 

13 Dec 447-12 Dec 446 


Artaxerxes I 20 

13 Dec 446-11 Dec 445 


Artaxerxes I 21 

12 Dec 445-11 Dec 444 


Artaxerxes I 22 

12 Dec 444-11 Dec 443 


Artaxerxes I 23 

12 Dec 443-11 Dec 442 


Artaxerxes I 24 

12 Dec 442-10 Dec 441 


Artaxerxes I 25 

11 Dec 441-10 Dec 440 


Artaxerxes I 26 

11 Dec 440-10 Dec 439 


Artaxerxes I 27 

11 Dec 439-10 Dec 438 


Artaxerxes I 28 

11 Dec 438-9 Dec 437 


Artaxerxes I 29 

10 Dec 437-9 Dec 436 

{continued on next page) 



Table (cont.) 

Era of 


Length of 365-day 


Regnal Year 

Egyptian Year 


Artaxerxes I 30 

10 Dec 436-9 Dec 435 


Artaxerxes I 31 

10 Dec 435-9 Dec 434 


Artaxerxes I 32 

10 Dec 434-8 Dec 433 


Artaxerxes I 33 

9 Dec 433-8 Dec 432 


Artaxerxes I 34 

9 Dec 432-8 Dec 431 


Artaxerxes I 35 

9 Dec 431-8 Dec 430 


Artaxerxes I 36 

9 Dec 430-7 Dec 429 


Artaxerxes I 37 

8 Dec 429-7 Dec 428 


Artaxerxes I 38 

8 Dec 428-7 Dec 427 


Artaxerxes I 39 

8 Dec 427-7 Dec 426 


Artaxerxes I 40 

8 Dec 426-6 Dec 425 


Artaxerxes I 41™ 

1 Dec 425- 
-death (Feb 424?) 

ephemeral rulers™ 

death Artaxerxes I— 
-6 Dec 424 


ephemeral rulers™ 

7 Dec 424-Darius IPs 
accession on Day x in 
25 Dec 424-13 Feb 423 

Darius IPs 

Darius IPs accession 

accession year" 1 

-ca. 1 1 Apr 423 


us II 1™ 

ca. 11 Apr 423-6 Dec 423 



us II 2 

7 Dec 423-6 Dec 422 



us II 3 

7 Dec 422-5 Dec 421 



us II 4 

6 Dec 421-5 Dec 420 



us II 5 

6 Dec 420-5 Dec 419 



us II 6 

6 Dec 419-5 Dec 418 



us II 7 

6 Dec 418-4 Dec 417 



us II 8 

5 Dec 417-4 Dec 416 



us II 9 

5 Dec 416-4 Dec 415 



us II 10 

5 Dec 415-4 Dec 414 



us II 11 

5 Dec 414-3 Dec 413 



us II 12 

4 Dec 413-3 Dec 412 



us II 13 

4 Dec 412-3 Dec 411 



us II 14 

4 Dec 411-3 Dec 410 



us II 15 

4 Dec 410-2 Dec 409 



us II 16 

3 Dec 409-2 Dec 408 



us II 17 

3 Dec 408-2 Dec 407 



us II 18 

3 Dec 407-2 Dec 406 


Darius II 19 

3 Dec 406- 

—at least until ca. 1 7 Sep 405™" 

Darius II 19 and 

ca. 18 Sep 405- 

Artaxerxes IPs accession year? 

-1 Dec 405" ;il 

(continued on next page) 



Table (cont.) 

Era of Ruler+ 

Nabonassar Regnal Year 

Length of 365-day 
Egyptian Year 

344 xx " 

Darius II 19 or 

Artaxerxes II's accession year'. 

Artaxerxes II 1 


Artaxerxes II 2 


Artaxerxes II 3 


Artaxerxes II 4 


Artaxerxes II 5 

2 Dec 405- 

-ca. 10 Apr 404^ 
ca. 10 Apr 404- 

-1 Dec 404 x: 
2 Dec 404-1 Dec 403 
2 Dec 403-1 Dec 402 
2 Dec 402-30 Nov 401 
1 Dec 401-30 Nov 400 

About this time, Persia lost control of Egypt.™' The native kings of Dyns. 
28—30 ruled for about seven decades. Their reigns are not day-exact, except 
Nectanebo IPs of Dyn. 30. The following provisional estimates (see Lloyd, 
CAH 2 , vol. 6, 358) are probably correct to within one or two years."™ 

Dynasty 28 
Dynasty 29 
Nepherites I 
Nepherites II 
Dynasty 30 
Nectanebo I i 


ca. 404/3-398/7 

ca. 398/7-392/1 
ca. 392/1-379/8 
brief reign 
brief reign 

ca. 379/8-361/0 
ca. 361/0-359/8 

The reign of the third king of Dyn. 30 and the last native ruler of Egypt, 
Nectanebo II (Nectanebos), is again day-exact, with high probability, as follows. 


Nectanebo II 1 


Nectanebo II 2 


Nectanebo II 3 


Nectanebo II 4 


Nectanebo II 5 


Nectanebo II 6 


Nectanebo II 7 


Nectanebo II 8 


Nectanebo II 9 


Nectanebo II 10 


Nectanebo II 1 1 


Nectanebo II 12 


Nectanebo II 13 


Nectanebo II 14 

21 Nov 359 

-20 Nov 358 : 

21 Nov 358- 

19 Nov 357 

20 Nov 357- 

-19 Nov 356 

20 Nov 356- 

-19 Nov 355 

20 Nov 355- 

-19 Nov 354 

20 Nov 354- 

18 Nov 353 

19 Nov 353- 

-18 Nov 352 

19 Nov 352- 

-18 Nov 351 

19 Nov 351- 

-18 Nov 350 

19 Nov 350- 

17 Nov 349 

18 Nov 349 

-17 Nov 348 

18 Nov 348- 

-17 Nov 347 

18 Nov 347- 

-17 Nov 346 

18 Nov 346-16 Nov 345 

(continued on next page) 



Table (cont.) 

Era of 


Regnal Year 

Length of 365-day 
Egyptian Year 


Nectanebo II 15 


Nectanebo II 16 


Nectanebo II 17 


Nectanebo II 18 


Nectanebo II 19 


Nectanebo II 20 

17 Nov 345-16 Nov 344 
17 Nov 344-16 Nov 343 
17 Nov 343-16 Nov 342 
17 Nov 342-15 Nov 341 
16 Nov 341-15 Nov 340" 
16 Nov 340-15 Nov 339" 

The following years are Egyptian years derived from the Canon (see section 3 


Artaxerxes III 20 


Artaxerxes III 21 


Arses 1 


Arses 2 


Darius III 1 


Darius III 2 


Darius III 3 


Darius III 4 


Alexander 1 

and so on 

16 Nov 340-15 Nov 339 
16 Nov 339-15 Nov 338 
16 Nov 338-14 Nov 337 
15 Nov 337-14 Nov 336 
15 Nov 336-14 Nov 335 
15 Nov 335-14 Nov 334 
15 Nov 334-13 Nov 333 
14 Nov 333-13 Nov 332 
14 Nov 332-13 Nov 331 
and so on 

1 Taharqa presumably came to the throne some day in this Egyptian year. There 
is a possibility that Year 1 is 12 Feb 691-11 Feb 690 (see section 2 above). No doc- 
uments dating to Taharqa's Years 1 and 2 are known (cf. Chapter III. 11, section 3). 

" Psammetichus I's reign presumably began in this Egyptian year. No documents 
dating to his Years 1-8 are known. 

'" The earliest date for Necho II is 19 Nov 610 BC, or Year 1 Month 11 Day 1, 
in stela Leyden V 18-9 (Kienitz, Geschichte, 155, with note 6, and 157-158). 

,v The latest date for Necho II is 4 May 595 BC, or Year 16 Month 4 Day 16, in 
stela Louvre 193 (Kienitz, Geschichte, 155, with note 2, and 158). The earliest date for 
Psammetichus II is 23 Nov 594 BC, or Year 1 Month 1 1 Day 9, in stela Louvre 240 
(Kienitz, Geschichte, 155, with note 4, and 158). 

v Psammetichus II died on 9 Feb 589 BC, or Year 7 Month 1 Day 23. The next 
day is here taken as Day 1 of Apries's reign. 

" The latest date for Apries is 19 Oct 570 BC, or Year 20 Month 10 Day 10, in 
abnormal hieratic pBM 10113 (M. Malinine, Choix de textes juridiques, I (Paris, 1953), 
17). The same document even anticipates a regnal Year 21 for Apries. The earliest 
date for Amasis is 1 1 JuL9 Aug 570 BC, or Year 1 Month 7, in a stela from the 
northwest Delta (Kienitz, Geschichte, 158, with note 4; G. Maspero, "Sur deux steles 
recemment decouvertes", RecTrav 15 (1893), 84-86, at 86; E. Edel, "Amasis und 
Nebukadnezar II", GM 29 (1978), 13-20, at 13). 

It has hitherto remained without explicit notice, as far as I know, that the latest 
date for Apries is later than the earliest date for Amasis, although S. P. Vleeming ("The 
Sale of a Slave in the Time of Pharaoh Py", OMRO 61 (1980), 1-17, at 6, note 20) 
rightly calls the Apries date "rather high." It now becomes possible to buttress the 


veracity of Herodotus's account (II, 169) about the overlap of the reigns of Apries and 
Amasis. After a short period with two rival Pharaohs, Amasis first vanquished Apries 
but then let him rule as coregent before eventually killing him. It is not clear whether 
Apries was a rival ruler or a nominal coregent on 19 Oct 570 BC, the date of pBM 
10113. It is significant that pBM 10113 is, like all abnormal hieratic texts, from Thebes, 
which was Apries's power base as rival ruler, whereas the Amasis stela mentioned 
above is from the Delta, Amasis's original power base. 

™ For the chronology of the transition from Amasis to Cambyses, see section 2 
above. Amasis could still have been in power on 1 Jan 526 BC as last day of his Year 
44. Greek sources mention a king Psammetichus, commonly styled as "III," who ruled 
six months between Amasis and Cambyses. The sole native documents ever assigned 
to his reign have recently been re-dated to a Psammetichus "IV," who may have ruled 
part of Egypt in the 480s BC (see E. Cruz-Uribe, "On the Existence of Psammetichus 
IV," Serapis 5 (1980), 35-39; cf. also P. W. Pestman, "The Diospolis Parva Documents: 
Chronological Problems concerning Psammetichus III and IV", in: Grammata Demotika, 
H.-J. Thissen & K.-Th. Zauzich, eds., (Wurzburg, 1984), 145-155; S. P. Vleeming, 
The Gooseherds of Hon (Pap. Hon) (Leuven: Studia demotica 3, 1991), 3-4). 

"" With this Egyptian year, Year 223 from Nabonassar, the present table joins 
Ptolemy's Canon, which earlier lists rulers of Babylon (see Chapter III. 11, section 
2.1). In the Canon, whose years are all full 365-day Egyptian years, Year 223 from 
Nabonassar is entirely Year 5 of Cambyses. 

" In Ptolemy's Canon, Year 226 from Nabonassar is entirely Year 8 of Cambyses 
and Year 227 entirely Year 1 of Darius I. 

x Cambyses was still recognized in Babylon in April 522 BC and, according to the 
Behistun inscription, did not die till after 1 Jul 522 BC (Parker & Dubberstein, Chronology, 

" The accession to the throne probably occurred between about 29 Sep 522 BC, 
the date of Bardiya's defeat according to the Behistun inscription, and about 22 Dec 
522 BC, the date of the earliest Babylonian tablet of Darius I's reign (Parker & 
Dubberstein, Chronology, 15). Again, lunar dates are mostiy not known to the exact 
Julian day but cannot be more than one to two days off. 

x " It is not known whether the time from the Egyptian new year of 1 Jan 521 BC 
to the Babylonian new year of about 13 Apr 521 BC was called Year 1 in Egypt or 
considered part of a numberless accession year (for an argument in support of the lat- 
ter, see L. Depuydt, "Regnal Years and Civil Calendar in Achaemenid Egypt;" JEA 
81 (1995), 151-173, at 164). In Babylon, for sure, regnal Year 1 did not begin before 
the first Babylonian new year of the reign. 

xl " In Ptolemy' Canon, Year 262 from Nabonassar is entirely Year 36 of Darius I 
and Year 263 entirely year 1 of Xerxes I. 

xlv The latest date for Darius I is about 24 Nov 486 BC (Parker & Dubberstein, 
Chronology, 17). The earliest known date for Xerxes I is about 1 Dec 486 BC (M.W. 
Stolper, "Babylonian Evidence for the End of the Reign of Darius I: A Correction", 
JNES 51 (1992), 61-62; cf. Depuydt, JEA 81, 157, note 22). 

xv It is not clear whether, in Egypt, the time from the Egyptian new year of 23 Dec 
486 BC, to the Babylonian new year of about 5 Apr 485 BC was called Year 1 or 
considered part of as numberless accession year. Cf. note xii above. 

™ In Ptolemy's Canon, Year 283 from Nabonassar is entirely Year 21 of Xerxes 
and Year 284 entirely Year 1 of Artaxerxes I. 

x ™ Xerxes I may have been murdered on one of these days. The source is an eclipse 
text (Parker & Dubberstein, Chronology, 17). 

x ™ This is the first day after the earliest possible date for the murder of Xerxes I 
(see note xvii). The earliest date for Artaxerxes I is 2 Jan 464 BC in an Aramaic 
papyrus ("B2.2" in B. Porten and A. Yardeni, Textbook of Aramaic Documents from Ancient 
(Jerusalem, 1986-99). Since the date is from Aswan, Artaxerxes I was probably 


in power already in 465 BC and perhaps even before the Egyptian new year of 16 
Dec 465 BC. His Babylonian regnal Year 1 certainly began about 13 Apr 464 BC. 
His Egyptian year may have begun earlier, but hardly before the Egyptian new year 
of 17 Dec 465 BC (cf notes xii and xix). 

" x It is not known whether the period from the Egyptian new year of 17 Dec 465 
BC to the Babylonian new year of about 1 3 Apr 464 BC was called Year 1 in Egypt 
or considered part of a numberless accession year (cf. notes xii and xviii). 

xx In Ptolemy's Canon, Year 324 from Nabonassar is entirely Year 41 of Artaxerxes 
I and Year 325 entirely Year 1 of Darius II. 

™ For problems pertaining to the chronology of the transition from Artaxerxes I to 
Darius II, see L. Depuydt, "The Date of Death of Artaxerxes I", WdO 26 (1995), 
86-96; id., JEA 81, 159, note 28. Arguments can be produced for the following pos- 
sible scenario: (1) Artaxerxes I dies in Feb 424 BC; (2) ephemeral kings including 
Xerxes II and Sogdianus rule for several months, while documents keep being dated 
according to Artaxerxes I, whose fictional Babylonian Year 41 began about 24 Apr 
424 BC; (3) Darius II assumes power some day in 25 Dec 424 BC-13 Feb 423 BC, 
perhaps rather near the end of this period. Darius II's Year 1 had certainly begun by 
about 10 April 404 BC, the Babylonian new year. 

xxn j n P( i enl y' s Canon, Year 343 from Nabonassar is entirely Year 19 of Darius II 
and Year 344 entirely Year 1 of Artaxerxes II. 

™" It is certain that Artaxerxes II's Babylonian regnal Year 1 began on about 10 
April 404 BC, the Babylonian new year. The latest date for Darius II is about 17 Sep 
405 BC, the date of Louvre cuneiform tablet AO 17603 (cf. Depuydt, JEA 81, 159, 
n. 29). It is not known when in the period from about 17 Sep 405 BC to about 10 
Apr 404 BC the transition from Darius II to Artaxerxes II happened. 

xxn -p^g i atest ,-i ate j s 18 Jan 401 BC (see Chapter III. 11, section 3). 

xxv Ptolemy's Canon continues at this point with regnal Years 6-46 of Artaxerxes 
II, measured by the Egyptian calendar. The New Year's days of these 41 full Egyptian 
civil years, Years 349 to 389 from Nabonassar, are as follows: 1 Dec 400-398 BC in 
years 349—51 from Nabonassar (= Artaxerxes II's regnal Years 6-8); 30 Nov 397-4 
BC in 352-5 Nab. (= 9-12); 29 Nov 393-0 BC in 356-9 Nab. (= 13-6); 28 Nov 
389-6 BC in 360-3 Nab. (= 17-20); 27 Nov 385-2 BC in 364-7 Nab. (21-4); 
26 Nov 381-78 BC in 368-71 Nab. (= 25-28); 25 Nov 377-4 BC in 372-5 Nab. 
(= 29-32); 24 Nov 373-0 BC in 376-9 Nab. (= 33-6); 23 Nov 369-6 BC in 380-3 
Nab. (= 37-40); 22 Nov 365-2 BC in 384-7 Nab. (= 41-4); 21 Nov 361-0 BC in 
388-9 Nab. (= 45-6). 

xxvl It is not known when Nectanebo II assumed power in this Egyptian year. 

xxvn , j/] lere j s no hieroglyphic evidence for Years 19 and 20 of Nectanebo II. 




Karola Zjbelius-Chen 

The era of the independent kingdom of Kush in Nubia is broadly 
divided into four parts: 

1. Pre-Dyn. 25, i.e. the epoch of the preliterate rulers of el-Kurru with 
approximately 6 reigns up to and including Alara, known from their 
tombs in the necropolis of el-Kurru (ca. 885/835 BC 1 to ca. 765 BC). 2 

2. Dyn. 25, simultaneously a segment of Egyptian history — and differing 
from Manetho and Egyptological tradition — plus the Kushite Kashta 
(Gen. 1) as its first king, since he secured the Thebaid by having 
his daughter Amenirdis I adopted by the Egyptian God's Wife 
Shepenupet I, daughter of Osorkon III. 3 Thus his daughter suc- 
ceeded Shepenupet in her official role at Thebes. The Assyrians 
effectively ended the dynasty, and Psammetichus I expelled the last 
Kushites at the time of Tanwetamani (ca. 655 BC). 

3. The Napatan Period, subdivided into an Early Napatan Period 
(Atlanersa (Gen. 7) to Malowi-Amani (Gen. 19), 4 ca. 653 until the 
mid-5th century BC) and the Late Napatan period, from Talakhamani 
and including the so-called Neo-Ramesside rulers with 5 kings up 
to Sabrakamani'' (mid 5th century to early 3rd century BC). 

1 See T. Kendall, Meroitica 15 (1999), 97. A long chronology (most recently L. Torok, 
Meroitica 14 (1999), 149ff.) no longer seems tenable after Kendall's article. 

2 Conventional dates reckoned back from the Tang-i Var inscription (see below) 
with the accession of Shebitku in 707: Shabaka, 15 years (highest known date: statue 
of Jty, BM 24429: Leclant, Enqwtes, 15ff); Piankhi, 31 years inferred; Kashta (= Manetho's 
Ammeris Aithiops), 12 years assumed; Alara, 20 years presumed; plus roughly 50 to 
1 00 years for the five reigns between Alara and the foundation of the el-Kurru necrop- 
olis. (For the reinstatement of the reading Piankhi, see C. Rilly, BIFAO 101 (2001), 35 Iff.) 

3 So R. G. Morkot, Meroitica 15 (1999), 195f, contra Kitchen, TIP\ § 122, who 
assumes that Amenirdis I was installed by her brother Piankhi. 

4 I prefer this transcription to Malowiebamani (Macadam, Kawa I, 54 (25), with 

«"> Jf reproducing Meroitic -b indicating a plural). I suppose instead that j -f~ is a 
phonetic determinative, derived from Egyptian w'j {Wb I, 246). 

5 Contrast Hofmann, Chronoloeie, 64ff. 


4. The Meroitic Period with the Early Meroitic period from Arkamaniqo/ 
Ergamenes I (Gen. 33) up to Queen Sanakadakhete (ca. early 3rd 
century until late 2nd century BC), the Middle Meroitic period (late 
2nd century BC to late 1st or early 2nd century AD), followed by 
the Late Meroitic Period (late 1st or early 2nd century AD until the 
end of the Meroitic kingdom). 6 

In the southern part of the kingdom, the post-Meroitic Period follows' 
with off-shoots of Meroitic culture surviving the end of the Kingdom 
between 330 and 370 AD (when the royal necropolis at Begerawiya 
North and the West cemetery at Begerawiya were finally abandoned) 8 
and the Ballana culture (so-called X-Group) in Lower Nubia with royal 
necropoleis at Qustul and Ballana which are beyond the chronological 
range concerning us here. 

This historical division into four periods ultimately depends on link- 
ing the seriation of the tombs in the royal necropoleis of the kingdom 
to other monuments and epigraphic data. Individual rulers are num- 
bered according to their relative positions in a scheme of generations 
(Gen.), especially since existing royal tombs cannot be assigned in every 
case to specific named rulers, and since some known rulers cannot yet 
be associated with any tomb. The cemeteries are el-Kurru (Ku.), Nuri 
(Nu.), Gebel Barkal (Bar.), Begerawiya South (Beg. S., with only two 
kings' tombs but other royal and non-royal burials), Begerawiya North 
(Beg. N.) and Begerawiya West (Beg. W., used by lesser members of 
the royal family and commoners). The excavator Reisner 9 established 
a typology, based on the architectural evolution of the tombs, and he 
identified tomb groups. 10 Reisner's archaeological seriation of the tombs, 

6 Torok, Meroe City — An ancient African capital: John Garstang's excavations in the Sudan 
(London, 1997), 13-14. 

' As viewed from a political perspective; see Torok, in: Welsby, Research, 142ff. 
Whether the cultural aspect of the transitional phase should be described as "post- 
pyramidal" (P. Lenoble, SARS Newsletter 3 (1992), 9ff.; idem, MM 25 (1994), 113f; 
idem, in: Welsby, Research, 157ff), remains open. 

8 The end of the Meroitic state and the abandonment of the royal cemetery in Beg. 
N. ca. 360/370 AD was approximately contemporaneous with the military campaign 
of the Axumite king Ezana against the Noba, unless Ezana's conversion to Christianity 
was earlier, around 330 AD. If so, the demise of Meroe will have been prolonged, from 
330 to 360/70 AD. See Torek's summary of the issues, in: Welsby, Research, 142ff. 

9 G. A. Reisner, "Preliminary Report on the Harvard-Boston excavations at Nuri: 
the kings of Ethiopia after Tirhaqa", HAS II (1918), 1-64; idem, SMR 2 (1919), 35-67, 
237-253; idem, JEA 9 (1923), 34-77, 157-160. 

10 Torok summarizes Reisner's criteria in: ANRW II 10, 169ff. 


which envisoned a continuous succession of 5 plus 68 rulers, 850 BC 
to 355 AD, 11 was revised by Dunham when he published the excava- 
tions. 12 Further revisions 13 were undertaken by Macadam 14 and, for the 
Meroitic material in particular, by Hintze 1 ' and Wenig. 16 The last also 
drew on the iconography of the tomb chapel reliefs which provides 
information about the tomb-owner (king, queen, prince), resulting in 
the elimination of some pyramid owners from the list of reigning mon- 
archs, but also on additional criteria for sequencing and thus dating 
tombs. Wenig's research eliminated Reisner's First Collateral Meroitic 
Dynasty of Napata, and integrated its rulers into the main line. The 
problem of the Second Collateral Dynasty of Napata, however, cannot 
be considered resolved. To include it, too, in the main line as Wenig 
proposes, seems justified, by demonstrable kinship relationships between 
pyramid owners in Barkal and Begerawiya North, aside from the fact 
that most of the rulers of the 1st century BC (to which the tombs 
belong) are attested both in the north and in Meroe. But if so, tem- 
porary changes in the royal burial grounds must be assumed, which 
could perhaps be explained in the context of the contemporaneous 
political situation. 17 

While Napatan internal chronology can generally be considered reli- 
able, 18 despite some specific problems and questions, 19 the same cannot 

11 The seriation, attempted by K. Grzymski, Meroitica 7 (1984), 222ff., using com- 
puter-aided quantitative analysis is generally considered unsuccessful. 

12 RCK I, II, IV, V. RCK IV, 6f.: 5 plus 72 rulers. 

13 On the typology for the last 20 generations, see F. Hinkel, Z^S 108 (1981), 105ff; 
idem, £AS 109 (1982), 27ff.; 127ff.; idem, Bulletin du Centenaire, Supplement BIFAO 81 
(1981), 379ff. 

14 Kawa I and II; idem, Allen Memorial Art Museum Bulletin 23 (1966), 42-71. 

15 Studien, 33 and Meroitica 1 (1973), 127-144: 5 plus 67 generations. 

16 MIO 13 (1967), 1-44 and Meroitica 1 (1973), 147-160: 5 plus 73 rulers; idem, 
Africa in Antiquity II (Brooklyn, 1978), 16f and Festschrift fur Steffen Wenig zum 65. Geburtstag 
(Niirnberger Blatter zur Archaologie. Sonderheft Sudan, 1999), 181-182: 6 plus 76 rulers. 

" Summary: Hintze, Meroitica 1 (1973), 135ff, with discussion of the conceivable 
historical scenario; see also Wenig, Meroitica 1 (1973), 152ff. 

18 Even in view of new data from the field, such as the finds at Gebel Barkal, with 
Bar. P-26 from the mid-6th century BC. This tomb is decorated with an astronomi- 
cal ceiling: F. Berenguer, Kush 17 (1997), 108ff; eadem, in: Welsby, Research, 55-63. 
Berenguer refers to P-26 as a royal tomb. It has only two chambers, yet contains a 
serekh with a name; there were no foundation deposits. 

19 The difficulties concern, e.g., the assignment of Nu. 20 and the position of 
Amanibakhi (Gen. 26) in the sequence; the stela from his chapel and an offering table 
were discovered in Nuri (RCK II, 269, fig. 213; R. J. Leprohon, CAA Boston Museum 
of Fine Arts 3, 127), but no tomb can yet be assigned to him. The sequence and relative 
chronology of the Neo-Ramesside rulers are also problematic, but according to their 


be said of the Meroitic material, where the situation is significantly 
more complex. In a fundamental review, Hofmann underscored the 
necessity of systematically utilizing Hellenistic and Roman imports in 
the tomb inventories for dating purposes, 20 but her work leads to a 
number of controversial problems, the most important being Hofmann's 
attribution of royal pyramids to persons whom she considers ruling 
kings. She bases her attibutions on offering tables with royal benedic- 
tion formulas K, L, and C, but the offering tables were not found in 
the royal necropolis itself. Furthermore, royal benediction formulas could 
also presumably have been used for members of the royal family. 21 
Torok discusses the chronological issues in the context of the publica- 
tion of the Fontes Historiae Nubiorum. 2 ' 1 A further difficulty in ascribing 
pyramids to known rulers ignores the fact that some pyramids were 
torn down (e.g., Beg. N. 53) or built over in antiquity. 

Wenig's proposal that stylistic elements be given more attention, and 
Hofmann's call for a more systematic examination of imports among 
the funeral deposits were virtually ignored, because the poorly preserved 
tomb chapels at Begerawiya have been inadequately explored archae- 
ologically. The challenges presented by interdisciplinary research essen- 
tial for dating the imports and evaluating their chronological and 
geographical distribution in the Mediterranean are also considerable. 
Nevertheless the results of the clearance, reconstruction, and docu- 
mentation of the pyramids of Meroe undertaken by Friedrich Hinkel 
and Janice Yellin and others, should eventually produce new evidence 
relevant to chronological issues. 23 

The relative chronology of the royal tombs, epigraphic sources, and 
other monuments furnish the foundations for establishing the sequence 
of the kings of Kush. Since the mechanism of succession remains unclear, 

titularies and the language of the texts (for which see now C. Peust, Das Napatanische: 
an dgyptischer Dialekt aus dan Nubien des spdten ersten vorchristlichai Jahrtausends (Gottingen, 
1999), 70-71) they belong at the end of the Napatan period. According to Morkot, 
in: Centuries of Darkness, P.James et al., eds., (London, 1991), 216f, the Neo-Ramesside 
rulers should be assigned to the period before Dyn. 25; more cautious idem, The Black 
Pharaohs (London, 2000), 146-150; see further below. 

20 Hofmann, Chronologie, passim (p. 192: 6 plus 69 rulers). 

21 Hintze, Studien, 62-63; Rilly, MM 28 (2001), 81. 

22 See the paragraphs for the reigns under discussion in FHN I III. 

23 Hinkel began studying the pyramids in 1976; there are 169 relief scenes at his 
disposal compared to 52 previously available; see Hinkel in: K. Bard, ed., Encyclopedia 
of the Archaeology of ancient Egypt (London, 1999,), s.v. Meroe; idem, Meroitica 7 (1984), 


and since succession from one generation to the next is paralleled by 
collateral succession among brothers and cousins, it is virtually impos- 
sible to project life spans and lengths of reigns from a change of sov- 
ereign. Genealogies established for individual sovereigns are largely 
based on circumstantial evidence or on conjecture about possible rules 
of succession; they are rarely certain and apply only to specific cases. 24 
For example, Assyrian sources designate Tanwetamani as the son both 
of Shabaka 25 and of Taharqa's sister. 26 His presumed mother is Qalhata, 
who should accordingly have been Taharqa's sister and Shabaka's con- 
sort. Her titles define her only as mwt nswt, snt nswt, hnwt n ti-stj and 
possibly z>t nswt (?), 2/ but not hmt nswt which is, however, not among 
the tides of any Kushite royal mother documented to date. Either these 
women were not royal wives, or the tide hmt nswt was deemed irrele- 
vant for them, possibly because it was considered a lower ranking tide, 
for even in those cases where the husband may have been king, he 
was dead at the time his son was proclaimed king. The reconstruction 
of kinship ties is rendered even more difficult since it is not known 
whether the Kushites employed the terminology in a literal sense, or 
whether terms such as snt "sister" and sn "brother" had broader con- 
notations. 28 For the period after Aspalta, who traced his maternal line 
back seven generations, 29 there is either very little genealogical or his- 
torical information, or none at all. The evidence and conjectural data 
suggest only that fully adult men ascended the throne. 30 

24 For a current, reliable study, see Morkot, Meroitica 15 (1999), 218-219, compar- 
ing idem, in the Preprint of the 7th International Conference for Meroitic Studies, Part 3, 
Appendix, 1992, 1-39 and especially Morkot's remarks on the royal succession in Dyn. 
25, Meroitica 15 (1999), 202-209. 

25 Egyptologists, except M. A. Leahy, CM 83 (1984), 43ff., consistendy consider 
Shabaka in error for Shebitku. However, it is improbable that the Assyrians erred in 
the filiation of their opponent, citing a king who had been dead for 43 years. 

26 See R. Borger, Beitrage zum Inschriftenwerk Assurbanipak (BIWA), 24 (A II 22 B II 
10) and 214f. 

27 A. Lohwasser, Meroitica 19 (2001), 180f. 

28 Cf. D. Apelt, Meroitica 12 (1990), 23ff. 

29 In his coronation stela: N. Grimal, Quatre steles napateennes au Musk du Caire: JE 
48863 et 48866 (Cairo: MIFAO 106, 1981), 30, 12-31, 3. The cartouches were all 
defaced. For one possible reconstruction and the problems it entails, see Morkot, Meroitica 
15 (1999), 196-200 and my review of Meroitica 15, 01^98 (2003), 441. 

30 Arike-Amanote, e.g., states that he was 41 years of age (Kawa LX, 1. 4). Tanwetamani, 
Shabaka's son, may have been in his mid-40's. If Taharqa left Nubia in 701 as a 20 
year old (Kawa V, 1. 20) to participate in the battle of Eltheke (K. A. Kitchen, in: 
Fontes atque pontes: eine Festgabe fir Helmut Brunner (Wiesbaden: AUAT 5, 1983, 249ff), 
then he may have been about 31 when he ascended the throne; Sabrakamani (Kawa 
XIII. 1. 2) gives his age as 39 (?). 


Similarly, it is not known whether the ruling sovereign influenced 
the nomination of a crown prince, if indeed such an institution existed 
(no source preserves a term for it). 31 Kawa Stelae IV, 7ff. and V, 13ff. 
state that Shebitku appreciated Taharqa more than any of the other 
royal "brothers", but this need not necessarily imply that Shebitku con- 
sidered Taharqa crown prince. 32 On the other hand, the Assyrian sources 
report that in 671 BC, Esarhaddon captured Taharqa's crown prince 
Ushanahuru. 33 But other Assyrian sources refer only to his son}* Kushite 
sources, and particularly the election stela of Aspalta, appear to testify 
against the practice of designating a successor. 

The only reign in more than twelve centuries of Kushite history 
whose length and absolute dates are known is Taharqa's (690-664 BC), 
fifth king of Dyn. 25 and eleventh (?) sovereign of Kush. The dates 
and lengths of his predecessors' reigns, especially those of his immedi- 
ate predecessor Shebitku (who had ascended the throne by 707 as the 
Tang-i Var inscription shows — see further below) and the others back 
to Alara (the first Kushite ruler whose name is known), 35 must be reck- 
oned from this point, as well as those of his successor Tanwetamani 
and the later Early Napatan rulers up to Aspalta. All other dates for 
Kushite rulers, particularly pre -Dyn. 25, but also those of the Napatan 
period after Aspalta, are estimates. There is a single certain date for 
the entire Meroitic period: 10 April 253 AD, in regnal year 3 of 
Teqorideamani (Beg. N. 28). 3(> All other year dates of this period are 
based on logical deductions from indirect testimony and historical con- 
siderations, frequently deriving from Classical sources. Some of these 

31 I doubt that hum. "youth" was such a designation (as Macadam suspected, Kawa 
I, 53 [6]). Taharqa, e.g., travelled to Egypt in the middle of the hwnw nfiw (plural), 
according to Kawa IV, 1. 8. The Adoption stela, 1. 4, describes Psammetichus II as jtj 

jw' Gb dmd psswy m hum, and even Amasis is hum nfr. Cf. also Zibelius-Chen, in: 
T. Kendall, ed., Nubian Studies 1998 (Boston, 2004), 468. 

32 As Macadam assumed, Kawa I, 17 (19). 

33 See R. Borger, Die Inschriften Asarhaddons, AfO Beih. 9 (1967), 99. 

34 In Chronicle 1 (A. K. Grayson, Assyrian and Babylonian Chronicles [Locust Valley, 
1975], 85) there is only a reference to his (i.e., Taharqa's) son and brother. See also 

J. Borker-Klahn, Altvorderasiatische Bildstelen (Mainz, 1982) Nos. 217. 218. Whether the 
Zincirli stela also depicts Ushanahuru is uncertain. According to Borker-Klahn, ibid., 
No. 219, the person wearing an uraeus could be Taharqa; however, Taharqa was not 
himself captured by Esarhaddon. 

30 For the nameless predecessors of Alara and their chronological classification, see 
Kendall, Meroitica 15 (1999), 3-117. 

36 Demotic graffito of Pasan (Ph. 416). 


estimates are well founded while others are conjectures. As a rule, 
neither the accession dates of Kushite kings nor their reign lengths are 
known, and only rarely can they be inferred. 

The reign of Taharqa (Gen. 5) depends upon the absolute dates of 
Dyn. 26 and those of its first king, Psammetichus I, whose first regnal 
year falls in 664 BC. 37 The stela Louvre 192 from the burial of an 
Apis bull in the Serapeum mentions that it was born in Taharqa's year 
26 and enthroned in the same year on 9/TV/peret; the bull died on 
20/TV/shemu in year 20 of Psammetichus I, having lived 2 1 years. In 
our calendar, the bull's death occurred on February 25, 644, and its 
enthronement on September 11, 665. Accordingly, Taharqa's year 26 
is the Egyptian year which lasted from February 6, 665 to February 
4, 664, when he died in the course of his 27th regnal year, having 
reigned 26 full years between 690 and 664 BC. 38 

The virtually certain sequence of Taharqa's predecessors in Dyn. 25 
is supported by the relative chronology of their pyramids, their inscrip- 
tions, and their representations. Assyrian synchronisms, together with 
the Tang-i Var inscription, show that Shebitku (Gen. 4) was already 
king in 707, or at the latest in 706. After Yamani of Ashdod fled by 
sea from the Assyrians in 711 and had dwelt "like a thief" in "the 
area of Egypt at the border by/to Kush", he must have been handed 
over to the Assyrian king Sargon II by the king of Meluhha/Kush in 
the year 706 at the latest. According to the Tang-i Var inscription and 
the version from Malatya, 39 this king is Shebitku. There is no evidence 
in favour of a supposed vice-royalty in Kush with Shebitku regent for 
Shabaka ruling in Egypt, nor for suggesting a coregency between the 
two kings, theorized to salvage the old chronology with its "anchor 
date" of 712 for the postulated campaign of Shabaka against or into 

" Beckerath, Chronologic, 41, 84-88. 

38 Beckerath, Chronologic, 91. According to Beckerath, SAK 29 (2001), 1 his first year 
began antedating on 12 February 690 BC. 

39 G. Frame, Orientalia 68 (1999), 40 [20]; N. Na'aman, JVABU 3 (1999), 63 [65]. 
For the chronological implications and interpretation of the historical/political events, 
see B. U. Schipper, Israel und Agypten in der Konigszeit (Fribourg, 1999, OBO 170), 200ff; 
K. Dallibor, Der antike Sudan 11 (2001), 41ff.; D. Kahn, Orientalia 70 (2001), Iff. and 
above all, A. Fuchs, Die Annalen des Jahres 711 v. Chr. nach Prismenftagmenten aus Ninive 
und Assur, The Neo- Assyrian Text Corpus Project (Helsinki 1998), 124ff. who reached 
the same conclusion about the date of Yamani's extradition before the Tang-i Var 
inscripton was made public. Note that D. B. Redford, Orientalia 68 (1999), 58ff. and 
K. A. Kitchen, in: Bietak, ed., SCIEM Haindorf 1996/98, 50f., as well as Beckerath, 
SAK 29 (2001), Iff, are sceptical of the value of these new sources. 


Egypt. 40 Consequently, D. Kahn rejects such ideas. 41 A division of power 
is documented neither for Piankhi nor Taharqa. That the Assyrians 
would have referred to the junior, less powerful ruler, is improbable. 
And why indeed should they not have mentioned both rulers of Egypt 
and Kush in their propaganda? It is more logical to suppose that 
Shebitku at his accession handed over Yamani, after he had spent some 
time in Egyptian-Kushite territory. Such an act might be viewed as a 
friendly gesture, since Yamani was an outlaw in Assyrian eyes, while 
in Egypt he could have stirred up trouble, perhaps seeking political 
support against the Assyrians. Both domestic and foreign policy could 
thus have provided motives for his extradition. Simultaneously, it might 
have served as a warning for the Delta princes to keep them in line. 
This scenario is, however, pure speculation, since neither the place of 
Yamani's sojourn nor what he did is known. Regardless, Shebitku did 
not want Yamani on his territory, and he failed to find asylum for him 
elsewhere, although committing himself to the anti-Assyrian alliance at 
Eltekeh shortly thereafter (701). Here, the petty rulers stood in battle 
array beside the Kushite army and their allies, as the Assyrian king 
Sennacherib specifically states; 42 the Delta rulers were thus still to be 
reckoned with, despite more than 30 years of Kushite rule in Egypt. 

Shebitku's reign can be calculated at a minimum of 17 years, 707—690. 
Africanus gives him 14 and Eusebius, 12. By contrast, his highest known 
date from Egypt is year 3. 43 His involvement in the Near East can only 
be grasped by studying Assyrian and Old Testament sources, as is also 
true for Taharqa. 44 

The accession of Taharqa's successor in 664 is certain. According to 
the Dream Stela 11. 6f., 45 Tanwetamani (Gen. 6) proceeded to Egypt 
in his first year, passing through Napata and Elephantine on the way 
to reclaim the Delta which had been lost to the Assyrians in 667. 
Indirect confirmation comes from Assyrian sources which refer to the 
change of rulers, noting that Tanwetamani made Thebes and Heliopolis 
his major bastions. Assyrian sources report that Tanwetamani was 

40 So, Beckerath, SAK 29 (2001), 4-5. 

41 Orientalia 70 (2001), 6-7. 

42 E. Frahm, Einleitung in die Sanherib-Inschriften, AfO Beiheft 26 (1997), 59. 

43 Nile Level mark No. 33 in Karnak; Kitchen, TIP", § 126. 

44 On Assurbanipal's battles against Taharqa and Tanwetamani, see the edition of 
textual sources by Borger (n. 26), 210-215. 

45 Grimal (n. 29), 7, 1 If. 


expelled from Thebes shortly thereafter (probably within only a mat- 
ter of months). 46 With the installation of Nitocris as God's Wife in 
Thebes by Psammetichus I ca. 655, Tanwetamani's rule in Upper Egypt 
ended. There is no hint of how long he continued to rule Kush. Using 
Greek sources, Burstein 47 plausibly argues that Tanwetamani did in fact 
return once to Memphis, where he fought Psammetichus I who had 
engaged the aid of Carian mercenaries. Tanwetamani was buried in 
the royal necropolis at Kurru, in Ku. 16. 

Psammetichus I fortified Elephantine and campaigned in LN, 48 but 
the identity of his Kushite opponent is not certain; chronologically, it 
could be either Atlanersa or Senkamanisken. 49 With the withdrawal of 
the Kushites from Egypt, Nubia sank into obscurity. Until the mid-4th 
century AD, which marks the end of the Meroitic kingdom, there were 
at least another 61 kings, but dated monuments are associated with 
few of them. Synchronisms with Egypt provide chronological fixpoints 
soley for the early Meroitic Period. Otherwise, there exists only Reisner's 
relative chronology, based on the typological criteria of the royal tombs 
and emended since only in some specific cases. Estimates of average 
reign lengths have been made, working with the number of kings 
between accepted fixpoints of the absolute chronology and by com- 
parison with Egyptian data of different periods; but these remain hypo- 
thetical. The proposed dates are merely suggested reference points 
indicating approximate positions, by contrast to Egyptian chronology 
for the LP, which is certain after 664 BC. 

The sequence of kings for the generations following Tanwetamani 
until Malonaqen (Gen. 12), based on archaeological seriation and inscrip- 
tions, is resonably reliable. The succession Atlanersa — Senkamanisken 
follows from the addition of the latter's name to the barque stand of 
Atlanersa (MFA 23.728). 50 Senkamanisken also completed Temple B 
700 at Barkal, begun by Atlanersa. Since the succeeding kings Anlamanf' 1 
and Aspalta 52 were both children of Nasalsa, it is clear that they were 

46 Borger (n. 26), 214 (B § 13, II 10-17). 

47 JSSEA 14 (1984), 3 Iff. 

48 L. Habachi, Oriens Antiquus 13 (1974), 323ff 

49 A fragment of an offering table belonging to Senkamanisken was found in Memphis: 
Zibelius-Chen, Meroitica 15 (1999), 712. 

50 Dunham, Barkal Temples, 32 (15); cf. also Reisner, Z^S 66 (1930), 91-92. 

51 Kawa VIII, representation in the lunette and 11. 22f. 

52 Stela of Madiqen from year 3, representation in the lunette, and on the stela of 
Khaliut. For the latter, see M.B. Reisner, %iS 70 (1934), 40 (13). 


brothers. Indirect evidence correlates Aspalta with Psammetichus II who 
campaigned against Nubia in 593, his 3rd regnal year.' 3 Traces of 
destruction at Napata which affected the monuments of Kushite rulers 
up to and including Aspalta (with later monuments untouched) are asso- 
ciated with Psammetichus's incursion. Ash layers were found in the cor- 
responding levels of Palace 1200 and in Temple B 500. Fires were 
likewise confirmed in the temple and "treasury" at Sanam, where the 
latest royal name found is Aspalta's. The chronological position of 
Aspalta thus inferred 54 was confirmed by the discovery of statues of 
Taharqa, Tanwetamani, Senkamanisken, Anlamani and Aspalta in a 
cachette in the temple precinct at Doukki Gel (Kerma), 55 all showing 
deliberate damage (including the removal of royal insignia). 

Generations 11 to 27, following Aspalta, span about three centuries, 
from the second quarter of the 6th century to the end of the 4th cen- 
tury BC (ca. 570—315 BC). Burials have been confirmed for 15 of the 
17 kings postulated. One tomb (Ku. 1), with a very large pyramid, is 
still unattributed, as is Ku. 2, presumably the burial of the consort or 
mother of Ku. l's owner. The temporary return from Nuri to the old 
royal necropolis at el-Kurru may reflect a brief change of dynasty. King 
Amanibakhi's mortuary stela and his offering table were discovered 
reused in Nuri 100, 36 but the king's pyramid has not yet been identified. 
He is tentatively placed before Nastasen (Gen. 21)? 1 Only a few royal 
monuments are known in this sequence before Nastasen, and those 
citing regnal years are quite rare. The highest known year for Aspalta 
is 3. For Arike-Amanote year 25+x is attested 58 and from Kawa 

oi For stelae of the king at Shellal, Karnak, and Tanis, see P. Der Manuelian, 
in the Past (London 1994), 337ff. Psammetichus II ascended the throne on 19/1/595 
BC (W. Barta, %AS 1 19, 1992, 89); his campaign in year 3 thus took place in 593 BC. 

54 Most recendy, Kendall, Rush 17 (1997), 232ff.; but cf. Torok, FHN I, 230-231, 
who considers the evidence insufficient for concluding that the Egyptian army reached 
and destroyed Napata. Torok thinks that the destruction could have resulted from a 
natural disaster or local political conflicts, in which case the survival of the Kdialiut 
stela undamaged and in situ is remarkable. Unfortunately it bears no date. Also undam- 
aged are the stela of Madiqen (dated to year 3; perhaps originally erected in Sanam), 
the inscriptions on the sphinx of Defeia, and the tomb and sarcophagus of Aspalta. 
Thus Aspelta could hardly have been the victim of persecution, as frequently claimed. 

55 For the find, see C. Bonnet, Genava 51 (2003), 267ff; D. Valbelle, RdE 54 (2003), 

56 RCKll, 269 fig. 213; for the stela, see also Leprohon (n. 19), 127ff. 

57 So now also Torok, FHN II, 465. 

58 Kawa XII. 


inscription IX, 4 it emerges that his predecessor was Talakhamani 
(Nu. 16). Thus the typological sequence postulated by Reisner has 
been confirmed by epigraphic evidence. 59 The mention of the ruler 
Malowi-Amani in Kawa IX, 12 demonstrates that he should be con- 
sidered the immediate predecessor of Talakhamani. 60 The annals of 
Harsijotef (Gen. 23) fall in his year 35, 61 and those of Nastasen (Gen. 
27) in his year 8. 62 

Archaeological remains confirm that after the campaigns of Psam- 
metichus I, LN was subject to Egyptian hegemony. Even under the 
Achaemenids, in Dyn. 27, the fort at Dorginarti remained for some 
time under northern control. Kushites appear as tribute bearers in the 
Apadana reliefs at Persepolis under Darius I, but their costume does 
not correspond to the Kushites' own imagery. On the statue of Darius 
I from Susa, ti-nhsj is listed among the occupied countries, along with 
Egypt and Libya. Cambyses is supposed to have attempted an inva- 
sion of (Upper) Nubia; and Xerxes deployed an Ethiopian contingent 
in his invasion of Greece. 63 The inscriptions of the Kushite kings 
Harsijotef and Nastasen reveal that they fought in LN, confirming that 
the region was then no longer subject to northern rule. Apparently, 
Kushite control extended northwards during the era of Egyptian inde- 
pendence from Persian rule, Dyns. 28—30 (404—342 BC). Harsijotef 
(Nu. 13) is thus placed around 404. It seems that Kushite kings also 
remained active in LN during the last phase of the Second Persian 
Period in Egypt, as Nastasen names Hmbswdn as the enemy in LN. 64 
Hintze, citing linguistic and historical arguments, equated this foe with 
the Egyptian rival king Khababash, 65 a proposal that strongly influenced 
the chronological position assigned to Nastasen and resulted in his being 
dated ca. 335-315 BC. 66 Although this identification of Nastasen's 

59 Macadam's proposal of a coregency between Talakhamani and Arike-Amanote 
was justly refuted by Vinogradov, Atti I. VI Congresso Intemazionale di Egittologia (Turin, 
1992/93), 635ff. 

60 So, too, Reisner HAS II (1919), 58. 

61 L. 1: Grimal (n. 29), 42, 2. 

62 L. 1: Peust (n. 19), 34. 

6:i L. A. Heidorn, in: W. V. Davies, ed., Egypt and Africa: Nubia from prehistory to Islam 
(London, 1991), 206ff.; Morkot, in: Achaemenid History Workshop 6 (Leiden, 1991), 324ff. 

64 L. 39; on the reading of the name, see most recendy, Peust (n. 19), 210-211. 

65 Studien, 17-20. 

66 On Khababash, who should probably be dated to 343-332 BC, see recently 
F. Hoffmann, Agypten (Berlin, 2000), 87 note 145. See also Peust (n. 19), 210-211. For 
the chronology of the Second Persian Period in Egypt, see D. Devauchelle, Transeuphratene 
10 (1995), 35-43 (non vidi). 


opponent (presumably a LN prince), is most assuredly inaccurate,'" the 
position of Nastasen in the sequence is unaffected. In terms of the gen- 
eral time scale, placing him somewhere in this era seems to be entirely 
reasonable, given the dates proposed for the early Meroitic king Arkamani- 
qo and the insertion of the so-called Neo-Ramesside rulers after Nastasen. 
The five "Neo-Ramesside" kings, Gen. 28-32, have been placed 
between Nastasen, the last king to be buried in Nuri (Nu. 15), and the 
first Meroitic king to be buried in Begerawiya South (Beg. S. 6). They 
are Aktisanes, 68 Ary(amani), 69 Kash . . . meri-Amun, 70 Arike-Pi(ankhi)-qo, 
and Sabrakamani. 71 But using archaeological criteria, it is possible to 
associate only three of the five with burials, viz. the anepigraphic Barkal 
pyramids Bar 11, 14, and 15, which belong between Nu. 15 and Beg. 
S. 6, the first royal burial at Meroe.' 2 Ascribing the pyramids to specific 
Neo-Ramesside kings is, however, completely hypothetical. Moreover, 
these kings are viewed as a group and placed in the same period solely 
on account of their names.'' 1 Their proximity to Amanislo (Gen. 34), 
buried in Beg. S. 5, and their position before Arnekhamani mrj-Jmn 
(Gen. 36), the builder of the Lion temple at Musawwarat es-Sufra may, 
however, be viewed as certain. Arnekhamani, buried in Beg. N. 53, 
was the last king to bear the epithet mrj Jmn (see below). 74 Amanislo 
and Amani-tekha 75 (Gen. 35) also used the epithet. However, it is not 
clear whether all five Neo-Ramesside kings form a single group in 
the period between Nastasen (Gen. 27) and Arkamani-qo (Gen. 33). 

67 Hintze identified the last part of the name as Meroitic wte (for vote, see now Rilly, 
MNL 27 [2000], 105 n. 5). It is, however, highly improbable that such a positive epi- 
thet would be attached to the name of an opponent; see A. S. Spalinger, ^AS 105 
(1978), 147. Furthermore, Khababash was chiefly active in LE. 

68 For Gtsn/Ktsn see K.-H. Priese, in: Fs Hintze, 343ff. Beckerath, HandbucK 1 , 275, 
"Atiaa-qo" is to be corrected; likewise, the mistaken data for H, G, and N names of 
Arike-Amanote (Beckerath, HandbucK 1 , 273). Diodorus I, 60 records an Ethiopian king 
Aktisanes, who might be identical with Gtsn/Ktsn. Diodorus took the name from Hekataios 
of Abdera, a contemporary of Ptolemy I, whose work must have been composed before 
300 BC; thus Aktisanes should be dated to before 300, or at the very latest, before 
282 BC. 

69 Kendall's identification as Alara (Meroitica 15 [1999], 64) cannot be maintained. 
'" Cf. Beckerath, HandbucK 1 , 275 Arkamanikasch . . . (?). 

71 All but Aktisanes are documented only in Kawa. See Macadam, Kama I, 72ff. 
(XIII, XIV, XV), 90 (XLV). 

72 See Wenig, MIO 13 (1967), Iff. 

' i I.e., the epithet mrj-Jmn written in the cartouche with the nomen of Kash . . ., 
Ary(amani) and Sabrakamani. 

'* Priese, in: Fs Hintze, 35 Iff. See also above, note 14. 

75 Probably restored correctly as Amani-tekha mrj-Jmn by Beckerath, HandbucK 1 , 277 . 


Peculiarities of writing and diction in the inscriptions of Sabrakamani 
and Arike-Pi(ankhi)-qo led Priese to date them earlier in the Napatan 
era, before Nastasen, 76 whereas Hofmann places Sabrakamani and 
Kash . . . amani in the Meroitic period (as Gen. 39 and 40), assigning 
them the later pyramids Bar. 7 and Bar. 9 in the north group at Gebel 
Barkal. 7 ' But this would imply that these two kings temporarily revived 
the tradition of using the epithet mrj-Jmn in their cartouches after a 
lapse of several generations. In the absence of additional arguments, 
this proposal does not seem very appealing. Positioning the Barkal pyra- 
mids 7 and 9 between the necropoleis of Nuri and Begerawiya led to 
the rejection of the idea of a First Collateral dynasty at Napata. 

Typologically, the tombs Beg. S. 6 und Beg. S. 5 at Meroe imme- 
diately follow the pyramids of the southern group at Barkal. They 
belong to kings Arkamani-qo (Gen. 33) and Amanislo (Gen. 34). According 
to Diodorus III, 6, the Ethiopian king Ergamenes was a contemporary 
of Ptolemy II (285-246 BC) and the first to oppose the priesthood and 
their custom of ritual regicide. This information resulted in the iden- 
tification of Ergamenes with Arkamani-qo, and thus to establish him 
approximately in the 2nd quarter of the 3rd century BC. The trans- 
fer of the royal necropolis from Barkal to Meroe in Arkamani-qo's 
reign also contributed to his identification with Ergamenes, as the move 
could have been related to the latter's opposition to the priesthood. 78 

After Amanislo, the sequence of royal tombs at Meroe continues 
in the northern area of Begerawiya. Of 41 tombs, at least two must 
be assigned to princes, and one to a non-reigning queen. The uncer- 
tainty of ascribing tombs in this cemetery results from the use of two- 
chambered tombs for kings after the turn of the 2nd to the 1 st century 
BC. Previously kings owned three-chamber tombs while two-chamber 
tombs were used for non-ruling members of the royal family. Since the 
non-ruling members continued to be buried in two-chamber tombs, it 
is difficult to identify the burial of a king when the reliefs in the tomb 
chapels have been destroyed, as is frequently the case; or when tomb 
stelae and offering tables are not preserved in situ. Furthermore, kings 
are usually not distinguished by the use of their title gore in the 

76 In: Fs Hintze, 352f. 

77 Chronologie, 65-66. 

78 F. Hintze, Die Inschriften des Lowentempels von Musawwarat Es Sufra (Berlin, 1962). 


nomination of the funerary texts. The series of tombs in the northern 
cemetery was interrupted for three generations by a move to Barkal, 
but then the sequence was resumed and maintained in the north ceme- 
tery until the end of the Meroitic kingdom. 

Associating individual pyramids with known kings to establish a 
chronological sequence is thus a particularly difficult task in this ceme- 
tery. The first tomb, Beg. N. 4, belongs to Amani-tekha who, distin- 
guished by the epithet mrj-Jmn in his cartouche, 79 follows Amanislo 
closely. The chronological position of Arnekhamani (Gen. 36), builder 
of the Lion Temple at Musawwarat es-Sufra, who has been assigned 
the destroyed pyramid Beg. N. 53, has been deduced from the repeated 
changes of epithet in his cartouche. Mrj-Jmn he first altered to c nh dt 
mrj-Jmn, and then to c nh dt mrj-Jst. This last epithet naming Isis is used 
in Egypt only by Ptolemy IV (221—205 BC). Presumably Arnekhamani 
ascended the throne during the reign of Ptolemy III (246—221 BC), 
and subsequently changed his name, when Ptolemy IV became pharaoh. 80 
Arqamani/Ergamenes II (Gen. 37; Beg. N. 7) and Adikhalamani/ 
Tabirqo? 81 (Gen. 38; Beg. N. 9) used the same epithet. They should 
therefore be successors of Arnekhamani, especially since Arqamani might 
be identical with the son of Arnekhamani who appears in the reliefs 
of the Lion Temple at Musawwarat es-Sufra. 82 A chronological niche 
for Arqamani and Adikhalamani is easily defined, since both built in 
the Dodekaschoinos in LN, 83 and can only have done so (for political 
and chronological reasons) during the Theban secession between 207/6 
and 186 BC. 

79 See above, note 75. 

80 Hintze (n. 78), 14f. The Egyptian rival king Hr-wn-nfr likewise takes the epithet 
'nh -dt mrj-Jst along with others at his coronation in Thebes, 205 BC, see W. HuB, 
Agypten in hellenistischer ^eit (Miinchen, 2001), 446 and P. W. Pestman, Chronologie egyp- 
tienne d'apres les textes demotiques (Leiden, 1967), 44. 

81 Possibly a mortuary name of Adikhalamani. 

82 Hintze, (n. 78), 25 (10). However, Hofmann, Chronologie, 57, rightly argues that 

^ -&&/!.*& ... 'liHuli Jl-ej-S 

Prince Arka l — i '^sHM£ cannot be Arqamani U A \»~~,T ^Jl^ ^- ; (Dakke; RCK IV, 

fig. D no. 24 G) because the orthography of the name is quite different, but she does 
concede that his chronological position after Arnekhamani could be correct. 

83 Chapel of Ergamenes in the temple of Dakke (G. Roeder, Der Tempel von Dakke, 
Cairo, 1911) and chapel of Adikhalamani in Debod (G. Roeder, Debod bis Kalabsche, Cairo, 
1911); see also a stela of the latter in Philae (A. Farid, MDAIK 34 [1978], 53-56). For 
the alleged joint constructions of Meroites and Ptolemies in the Dodekaschoinos, see 
Torok, Handbook, 210-211, 428-431. 


The tombs continuing the archaeological sequence in the northern 
cemetery can be assigned to specific individuals only with difficulty. 
Relevant epigraphic material is generally lacking, and attributions on 
the basis of fragmentary and displaced offering tables can be no more 
than provisional. A ruling queen must have been interred in the pyra- 
mid Beg. N. 1 1 (Gen. 41), where the reliefs depict a woman with royal 
insignia. Shards with Demotic and Meroitic cursive characters have 
been found in the debris, supporting attribution of it to Sanakadakhete. 
The earliest known inscription in Meroitic comes from Temple F which 
she built in Naqa. An iconographic detail leads Hofmann to correlate 
the pyramid with the later part (after 1 45 BC) of the reign of Ptolemy 
VIII (170-163, 145-131, and again 127-116 BC). 84 A block from a 
pylon with traces of a Horus name with the component k?-nht comes 
from Beg. N. 20 (Gen. 44). Ptolemy IX (116-107 and 88-80 BC) and 
Ptolemy XII (80—57 and 55—51 BC) use U-nht in their Horus names, 
suggesting to Wenig that the pyramid Beg. N. 20 belongs to the Meroitic 
king Taneyidamani, who is known from many monuments and must 
belong to this period." 5 However, Hofmann and Torok consider him 
the immediate successor of Sanakadakhete, and therefore assign him 
pyramid Beg. N. 12 (Gen. 42). 86 

After Beg. N. 20, the royal family moved the burial ground to Barkal 
for three generations, before returning to Begerawiya North for the 
remainder of Meroitic history. Only one of the three tombs at the lat- 
ter site can be assigned, viz. Bar. 6 to the ruling queen Nawidemaka. 
Other identifications depend upon differing interpretations of the his- 
torical events that may have led to the transfer of the necropolis. 87 
Since the assignment of the tomb group Beg. N. 11, 12, 13, and 20 
is already hypothetical, and the sequence of kings accordingly contested, 
any interpretations amount to mere circular reasoning until new mate- 
rial comes to light. 

Epigraphic documentation for three personalities — king Teriteqase, 
Kandake Amanirenase and Paqara Akinidada — at the temple of Dakke 88 

84 Chronologic, 11-1%. 

85 MIO 13 (1967), 43. 

86 Hofmann, Chronologie, 78. Torok notes palaeographic details favouring Taneyidamani 
for Sanakadakliete's immediate successor, Handbook, 205; idem, FUN II, 664. 

8/ But not, in my opinion, an active northern policy (cf. Hofmann, Chronologie, 66ff), 
since the kings of Dyn. 25 had a decided northern policy with the seat of government 
in Egypt, yet were buried in Kush. 

88 REM 0092; REM 0093. 


is chronologically very significant. Queen Amanirenase, who apparently 
reigned after the death of Teriteqase, is attested together with Akinidada 
in Kawa and in Hamadab. 89 Akinidada may be the son of the one- 
eyed Kandake mentioned by Strabo. 90 Akinidada is also named with 
the reigning queen Amanisakheto (Beg. N. 6) on a stela in Q_asr Ibrim. 91 
These four individuals must have been not only contemporaries, but 
also apparently associated with historical events linked to a more aggres- 
sive northern policy, which led to the Meroitic conquest and pillage of 
Philae and Syene and provoked the conflict with the Romans, resulting 
in Petronius's campaign of 23 BC. With the peace of Samos 21/20 BC, 
both parties were satisfied with the Roman garrison's withdrawal from 
Qasr Ibrim while the Meroites recognized the border at Maharraqa. 92 
Only in 297 AD did Diocletian move the frontier back to Syene, but 
exactly how the Romans ruled LN during their years of sovereignty, 
and the degree to which the Meroites may have been involved, can- 
not be established with certainty. Demotic graffiti in the Dodekaschoinos, 
dating for the most part to the 1st and 3rd centuries AD, testify to the 
activities of Meroitic officials in the region. The cult of Isis at Philae 
was particularly important to them, and defensive actions against the 
aggressive and marauding Blemmyes will have been a common con- 
cern of Romans and Meroites. The Kharamadoye text in the Mandulis 
temple at Kalabsha is the last Meroitic inscription in the region; 93 it 
dates to the beginning of the 5th century AD. 94 Whether King Aqraka- 
mani (>qrg-imn?) should also be assigned to the period of conflict with 
Petronius remains unclear. 95 A Demotic inscription of a Meroite in 
Dakke belongs to his reign and that of his royal mother, the pr-H Naytal 

mv 6 

89 REM 0628; REM 1039. 

90 Strabo 17, 1. 53f.; see FHNlll, 828ff. (190). 

91 REM 1141; cf. also REM 0705. 0706 in the temple T at Kawa. 

92 Summary of the historical events according to Roman sources: Welsby, The Kingdom 
of Kush: the Napatan and Meroitic Empires (London, 1996) 68ff.; Torok, Handbook, 448ff. 

93 For the dates, see Torok, FHNlll, 1105f. 

94 That the Kharamadoye inscription might have been written in a language other 
than Meroitic (as Peust (n. 19), 75, maintains) is highly improbable. It is precisely post- 
positions such as -dik or -lw (which Peust cites as indicators of a different language) 
which are known Meroitic language morphemes, like the lexeme qore "king". A glance 
at the beginning of the inscription, which clearly reveals Meroitic diction as far as line 
9, should have discouraged him from pursuing this proposal. For analyses of the inscrip- 
tion, see Bibliographie REM 0094 and Millet, MNL 30 (2003), 57-72. 

95 Torok in: FHN II, 686 (161), without, however, going into A. Burckhardt's argu- 
ments in Meroitica 8 (1985), 76 and n. 12, where this graffito is assigned to the late 
1st or early 2nd century AD. 

96 Dak. 17; cf. Burckhardt, Meroitica 8 (1985), 99. 


After Amanisakheto come the great Meroitic builders Natakamani 
(Beg. N. 22) and Amanitore (Beg. N. 1) under whose reign a "Re- 
Egyptianization" in art and script occurred. The new excavations of 
the Berlin Egyptian Museum discovered two stelae of Amanisakheto in 
the temple of Amun at Naqa, 9 ' built by Natakamani and Amanitore, 
which may necessitate a reassessment of the sequence of these rulers, 
unless the two stelae actually come from an earlier building in Naqa, 9K 
or were perhaps shifted to the Amun temple and left there. Both ste- 
lae show the queen with Amesemi, once before Amesemi's consort 
Apedemak (REM 1293), and once simply together (REM 1294). Studying 
details of the representations and the finds in the tombs, Hofmann 99 
dates Natakamani to the 2nd half of the 1st century AD. 

It is uncertain, although entirely plausible, that Sorakarora, who occa- 
sionally appears together with Natakamani and Amanitore and may 
have initially occupied the position of a pqr(tr), should be placed in the 
sequence of rulers after Natakamani. A relief at Gebel Qeili shows 
Sorakarora with full royal insignia. 100 

Links with the Roman Empire are lacking for the following epoch, 
and there are only a few kings attested in inscriptions and by offering 
tables. The small number of names which belong to known rulers, and 
those which have been proposed as names of rulers, are mere compi- 
lations in list form. Only the pyramids Beg. N. 17, 18, and 19 101 and 
the late pyramid Beg. N. 28 (of Teqorideamani) can be associated by 
inscriptions with owners and their sequence established beyond doubt, 
following Beg. N. 16. Beg. N. 17, 18, and 19 date to the period from 
somewhat after the mid- 1st century AD to the mid-2nd century. How- 
ever, two offering tables were found in Beg. N. 16, one belonging to 

97 REM 1293 and 1294. 

98 Naqa may already have been documented in the stela of Nastasen as trrqt, see 
Zibelius, Afrikanische Oris- und ViJlkernamen in hieroglyphischen und hieratischenTexten (Wiesbaden, 
1972), s.v. twrkt(t) and trrqt. 

99 Chronologic, 128ff.; see also Torok, Handbook, 46 Iff.; idem, FHN III, 898f. 

100 M. Zach (GM 136 [1993], 89ff.) considers Sorakarora a rival of King Natakamani; 
Hofmann (Chronologie, 128) removes him from the list of kings and assigns Beg. N. 15 
to him; Torok (FFLN III, 9 1 Of.) considers him a ruling king since he wears a royal cos- 
tume at Gebel Qeili; Zach (GM 145, 1995, 105ff.) concurs. See also Wenig (n. 16), 
17; idem, Fs Wenig (n. 16), 182. 

101 Amanitenamomide (Beg. N. 17), Queen Amanikhatashan (Egyptian hieroglyphs, 
Beg. N. 18), Tarekeniwala (Beg. N. 19, in the debris of which was found an offering 
table of Ariteneyesebokhe). Beg. N. 29 probably Takideamani, and certainly Teqoridea- 
mani (Beg. N. 18). 


Amanitaraqide 102 and the other to Arayesebokhe. Palaeographic and 
typological features allow earlier pyramid Beg. N. 16 to be ascribed to 
Amanitaraqide. Arayesebokhe is dated to a later period, since his offering 
table is typologically younger, and Beg. N. 36 is assigned to him, but 
there are significant discrepancies in his chronological position. 103 Further- 
more, it is by no means clear that both these individuals were kings 
at all: no other monuments have yet surfaced for either. 104 The assign- 
ment of pyramids in the following period also remains entirely hypo- 
thetical. For the specific identification of the burials in Hofmann's 6th 
group of pyramids in Begerawiya North, which she dates from the mid- 
2nd century to the mid-3rd century, she relies on offering tables with 
royal benediction formulae found in the West necropolis at Begera- 
wiya, on the assumption that they had been moved. 103 This clearly 
illustrates the difficulties and uncertainties surrounding the identifications 
of rulers and their relative chronological positions. 106 

King Amanikhareqerema, who left us the "omphalos" (i.e., a naos) of 
Napata, 107 two ram statues 108 and a rounded stone with his name found 
at Naqa, 109 has as yet no tomb at all, 110 unless Beg. N. 37 (Gen. 67) U1 

102 His parents are Mnhdoke (B) and Pisekr (C). The facsimile REM 0816 has Pisekdo 
instead (C), but it is not possible to decide which reading is correct on the basis of 
the photograph. Pisekara appears in the king lists because kings' fathers are presumed 
to have been kings too, but this is by no means certain. 

103 p or example, Torok, FHN III, 9 12—9 14 places Arayesebokhe as Gen. 56 with 
Beg. N. 36 immediately after Amanitaraqide with Beg. N. 16. The same sequence 
appears in Torok, Handbook, 205, but with Beg. N. 36 assigned to Amanitaraqide and 
Beg. N. 16 to Arayesebokhe. Hintze, Studien, 33 placed Amanitaraqide in Gen. 48 and 
Arayesebokhe much later in Gen. 60, while Wenig (n. 16), 17; idem, Fs Wenig (n. 16), 
182) assumes Gens. 56 and 69, and Hofmann, Chronologie, 192, assigns them, respec- 
tively, to Gens. 52 and 63. 

104 See Rilly, MNL 28 (2001), 81. 

105 Chronologie, 155. 

me p or exam ple ; inserting Maloqorebara (Meroitic chamber in Philae, REM 0101 1. 
1) in the king list is fraught with difficulties; cf Torok' s discussion, FHJV III, 1028f. 
Even if a royal name should be expected in REM 0101 (so Torok) the context of the 
inscription (tdhe:Mloqorebr:qoretlhidemni . . . .) does not suggest that Maloqorebara should 
be considered a king. At our present understanding of the language qoretlhidemni 
/ qoretalahidemani/ is to be read as qorese-l Ihidemni. 

107 REM 1004. 

108 REM 0001 and REM 1151. 

109 REM 1282, with the writing Mnhereqerem {contra Rilly, MNL 28 [2001], 71 n. 1), 
see the figure, K. Kroeper & D. Wildung, ANM 9 (2002), 146 pi. VI a (inverted); 
G. & J. Hallof, B Z S 1 (2000), 169ff; C. Carrier, MNL 27 (2000), 2 and figs. 4-5. 

110 Torok, Handbook, 206 (Gen. 62). 

111 Wenig (n. 16), 17; idem, Fs Wenig (n. 16), 182. 


or Beg. N. 41 (Gen. 56) is his. 112 Proposed dates range from the 
2nd half to the end of the 2nd century. When doubts arose about the 
late dates for Amanikhareqerema, Rilly attempted to check the chrono- 
logical position proposed for him by applying palaeographic criteria to 
the inscription on the stone REM 1282. 113 To avoid circular reasoning, 
he relied on statistical analyses of the cursive royal inscriptions after 
Natakamani. His study includes 18 documents, analysed according to 
strict methodological criteria, while excluding archaeological and icono- 
graphic factors. Rilly concludes that the later dates for Amanikhareqerema 
should be reconsidered, since he may belong instead to the end of the 
1st century. 

We have almost no names for the kings reigning between Amani- 
khareqerema and those of around a century later. Many princes and 
queens were included in the king-list for the last centuries of the king- 
dom of Meroe, none of whom reigned autonomously. In the corpus 
studied by Rilly they are: Amanitaraqide, Arayesebokhe, Amanikhedolo, 
Mashaqadakhela, Temelordeamani, who may have been a half-brother 
of King Teqorideamani, Pat..rapeamani and Amanipilade. 

A maximum list of kings after Natakamani (and Sorakarora?) would 
include 23 rulers with a total of 24 pyramids. In a minimal list, only 
the following individuals can be considered kings: 

Amanikhareqerema, end 1st century AD 

Amanitenamomide (Beg. N. 17) 

Amanikhatashan (Beg. N. 18) 

Takideamani (?) 

Tarekeniwala (Beg. N. 19) 


Teqorideamani (Beg. N. 28), accession 248/249 AD 


Teqorideamani can be dated precisely; he is named in the Demotic 
graffito Ph. 4 1 6 of Pasan mentioning the third year of the ruler Tqrrmn 
and year 253 (in our reckoning) under Trebonius Gallus. The king's 
name, written in Meroitic Teqorideamani, was found in his pyramid, 
Beg. N. 28. We also possess an offering table REM 0829, dedicatory 
texts REM 0408 through REM 0410, and the graffito REM 1261 in 

112 Hofmann, Chronologic, 160. 

113 MNL 28 (2001), 7 Iff. 


Meroe inscribed with his name. He may have reigned 20 years. 114 The 
inscription of his official Pasan and those of other Meroites attest that 
the personnel of the Isis temple was partially under the authority of 
Meroitic officials. It is probably in this connection that the represen- 
tations in the Meroitic chamber at Philae showing a Meroitic delega- 
tion should be understood. 115 One of the successors of Teqorideamani 
was Yesebokheamani, who is attested in a dedicatory inscription on 
Hadrian's gate at Philae, where he is explicitely titled qore. ne He thus 
belongs most probably to the period of Roman withdrawal from the 
Dodekaschoinos. 117 His lion statue was found in Qasr Ibrim in LN, 118 
and he is also known from a stela in the Apedemak temple at Meroe, 
with an inscription identifying him as king. 119 

The sequence of pyramids in Begerawiya North ends with Beg. N. 
25. This tomb belonged to a queen who is depicted wearing royal 
insignia and sitting on a lion throne to receive an offering of incense. 120 
The pyramid Beg. N. 25 represents the last burial in the royal necrop- 
olis at Begerawiya North, marking the end of the Meroitic kingdom. 

114 Deduced by Hofmann, Chronologie, 168, based on Ph. 68, where year 20 of a 
Meroitic king is mentioned. 

115 REM 0097-0111. 

116 REM 0119 and REM 0120. On qoro < qore-lo see Rilly, MJVL 26 (1999), 79ff. 

117 See Torok, FHNlll, 1050; contrast Hofmann, Chronologie, 189. 

118 Not in REM. See J. M. Plumley, JEA 52 (1966), 12, pi. 4, 3. The writing can- 
not be confirmed on the basis of the published photograph, but see Hallof, JEA 89 
(2003), pi. 23. 

119 REM 0407, 1. 2 without the adjunct amni only as yesebohe:qore. 

120 RCK III, pi. 23. 




Jorg Klinger 

The multiple sources available for the reconstruction of the chrono- 
logical background of ancient Near Eastern history are a major factor 
distinguishing these cultures from Egypt, where ultimately there is but 
one single historical thread. By contrast, Mesopotamia and the neigh- 
bouring regions offer a series of at least partially independent sources. 
There are admittedly important differences in the distribution of these 
sources over time and space such that phases with abundant material 
contrast with others which are less satisfying. For certain periods, there 
is simply insufficient material to draw any conclusions, regardless of the 
number of different threads available. 

The middle of the second millennium is such a period, and this has 
led to the conundrum of three parallel chronological systems, each 
based on the Venus tablet recording year 6 of Ammisaduqa of Babylon. 1 
These Venus observations seem to have offered the basis for the astro- 
nomical calculations, since this observational event repeats itself every 
56/64 years, and thus the date means that the year 6 of Ammisaduqa 
can be set on a spectrum ranging from 1641 to 1577 BC. This fix 
point became a pivot upon which the various chronologies turned, 2 

1 The texts are part of the omina series Enuma Anu Enlil; cf. esp. E. Reiner & 
D. Pingree, Babylonian Planetary Omens. Part I: The Venus Tablet of Ammisaduqa (Malibu, 
1975). For an evaluation of the data, and the link with the lunar dates in particular, 
cf. also, V. G. Gurzadyan, "On the Astronomical Records and Babylonian Chronology", 
Akkadica 119-120 (2000), 180ff. and the literature listed there. 

2 Fundamental for establishing the chronological frameworks were the work of 
F. Cornelius, "Die Chronologie des Vorderen Orients im 2. Jahrtausend v. Chr.", AfO 
17 (1954-56), 294ff, with a short survey of the state of the chronological discussion 
up to that time, and a plea for the "Low Chronology", along with B. Landsberger, 
"Assyrische Konigsliste und 'Dunkles Zeitalter'", JCS 8 (1954), 31-45, 47-73, 106-133 
and A. Goetze, "On the Chronology of the Second Millennium BC", JCS 11 (1957), 
53-61, 63-73, favouring the "Middle Chronology", along with the response from 
F. Cornelius, "Chronology. Eine Erwiderung", JCS 12 (1958), 101-104. A standard 
table with dates for Mesopotamian rulers following the Middle Chronology was pre- 
pared by J. A. Brinkman for A. L. Oppenheim & E. Reiner, Ancient Mesopotamia. Portrait 
of a Dead Civilization (Chicago, 1977), 335-348. 


with the later periods firmly anchored, and the earlier periods floating 
freely, the date for the fall of Babylon being 1595 (according to the 
Middle chronology) or 1531 BC (according to the Low chronology). 3 
This point has now been lost since the observations are no longer 
viewed as reliable, 4 with the result that the long debate about the rel- 
ative merits of the Middle or Low chronologies must now be aban- 
doned, having become pointless as the various alternatives for an absolute 
date are no longer separated by intervals of 64 years. Instead, we now 
have a moving link separating a block of 500 years 5 from the later 
periods where the chronology is relatively reliable. b 

3 The conquest of Babylon by Murshili I is an important event in the Hittite ver- 
sions, but cannot be used chronologically. The Babylonian records indicate that it took 
place at the "time of Samsuditana"; Agum (II?) recovered of the plundered statues 
only 24 years later. The relevant references have been collected by G. Wilhelm, 
"Murshili", RIA VIII: 434-435; on the chronological evaluation, cf. A. Goetze, JCS 
11, 65-73, and for the difficulties of the Kassite kings named Agum, cf. esp. J. A. 
Brinkman, Materials and Studies for Kassite History (Chicago, 1976), I, 95, 97. 

1 Cf. most recendy, L. Sassmannshausen, "Babylonian Chronology of the 2nd Half 
of the 2nd Millennium BC", in: H. Hunger & R. Pruzsinszky, eds., Mesopotamian Dark 
Age Revisited. Proceedings of an International Conference ofSCIEM 2000, Vienna 8th-9th November 
2002 (Vienna, 2004), 65. 

On the evaluation of the dates, cf. P. J. Huber, "Astronomy and Ancient Chronology", 
Akkadka 119—120 (2000), 160—174, who still maintains that a statistical analysis of the 
possibly faulty dates is possible, and that the ensuing results which he considers to be 
compatible with a date for Ammisaduqa 1 are 1582 or 1516. 

5 The preservation of the very different strands from the beginning of the Ur III 
period (2111-2003 BC, Middle Chronology) to the end of the Old Babylonian period 
are so tightly woven that the relative chronology for the history of the first half of the 
second millennium can be viewed as certain; the prospective of either larger hiati or 
a significant reduction can be excluded. For the Ur III period, with its rich harvest of 
chronological data, cf. W. Sallaberger "Ur III-Zeit", in: W. Sallaberger & A. Westenholz, 
Mesopotamien: Akkade-^eit und Ur Illicit (Fribourg, 1999), esp. 123ff. 

6 Explicitly: the use of a date 1531 or 1595 BC for the Hittite conquest of Babylon, 
which is certainly relevant for the beginning of Hittite history, is today a mere con- 
vention; limiting it to a period of 10-20 years is possible. Cf. the basic observation by 
C. Kuhne, "Imperial Mittani: An Attempt at Historical Reconstruction", SCCMH 10 
(1999), 203 n. 1, who in the meantime assumes a decade between "1550 and 1540 
as a transitional decade for the end of the Old Babylonian Dynasty". Referring to the 
running dendrochronological analyses by P. I. Kuniholm (cf. also S. W. Manning, 
B. Kromer, P. I. Kuniholm & M. W. Newton, "Anatolian Tree Rings and a New 
Chronology for the East Mediterranean Bronze-Iron Ages", Science 294 (2001), 2532—35; 
M. W. Newton & P. I. Kuniholm, "A Dendrochronological Framework for the Assyrian 
Colony Period in Asia Minor", Tiirkiye Bilimler Akademisi Arkeoloji Dergisi 1 (2004), 165ff. 
and the current reports of the Aegean Dendrochronology Project at http://www.arts.cor-, he proposes 1541 BC +/- 37 years as a possible date for the Babylon 


It was just a question of time until a new attempt would be made 
to revise the entire chronology of the period before the first millen- 
nium. The model proposed by Gasche et al. in 1998' thus gave momen- 
tum to a new and intense debate which endures up to the time of 
writing." Whether a new chronological framework similar to that pre- 
vailing with the Middle Chronology will emerge remains to be seen. 
New sources are still not available, with the exception of the gradual 
clarification of the royal succession in the Middle Hittite period, which 
can at least offer some kind of orientation around the middle of the 
second millennium. 9 At present, the earlier chronological proposals are 
maintained as conventions. 

The greatest lack in the cuneiform sources thus continues to be the 
lack of adequate data 10 combined with the absence of an overarching 
system of year names." 

In Assyria, the traditional principle of naming years and counting 
according to the holder of a certain office (limti) was maintained, but 
the sequence of officials in the books of eponyms eliminated the need 
for naming the years. This system was already established in the first 
quarter of the second millennium, and became a standard dating method, 
as can be seen in the eponym dates in the thousands of Old Assyrian 
texts from the karum-period in Anatolia, and as is confirmed by the 
recent discovery of two tablets with limu lists. 12 With the overlap between 

7 Gasche et al., Dating. 

8 Thus a recent attempt at an ultra-long chronology, cf. C. Eder, "Assyrische 
Distanzangaben und die absolute Chronologie Vorderasiens", AoF 31 (2004), 191—236. 

9 There are a whole series of astronomical observations completely independent of 
these which are highly significant and could contribute to gaining a higher resolution 
for the DeltaT (Deceleration of the rotation of the earth) factor which is so important 
and difficult in these calculations. In addition, there are specific cases where chrono- 
logically absolute dates can be checked or controlled using calculable dates. Here, we 
will merely refer to the detailed investigations by Stephenson, Eclipses. Specifically rel- 
evant for the Ancient Near East are chapters 4-7; cf. also below Locher, chapter III. 9. 

10 Cf. A. Ungnad, "Datenlisten", RIA II: 131-194 and Oppenheim (n. 2), 145-146. 

11 An illustration of selection of such year names can be found in translation in 
W.H.Ph. Rbmer, in: THAT I, Fasc. 4, (Giitersloh, 1984), 337ff. For the year names, 
cf, above all, the relevant article by A. Ungnad, "Eponymen", RIA II: 412-457, and 
the compilation of the literature with possible additions in Gasche et al., Dating, 47 
n. 200 and the short survey in Sallaberger (n. 5), 231-237. 

12 On the discovery of the lists and an initial evaluation of their chronological rel- 
evance, cf. K. R. Veenhof, "Old Assyrian Chronology", Akkadita 119-120 (2000), 137-150; 
these are now published: K. R. Veenhof, The Old Assyrian List of Year Eponyms from Karum 
Kanish and its Chronological Implications, Ataturk Supreme Council for Culture, Language 
and History — Publications of the Turkish Historical Society Serial VI — No. 64 (Ankara, 
2003. The discovery of additional texts of this genre has since been announced. 


these lists and the eponym chronicle from Mari, a period of almost 
exacdy 200 years is not absolutely precisely dated in relative terms. 
This allows the end of karum Kanesh II to be placed into a relative 
chronological context, and this permits further deductions about the 
overall chronological framework of the whole period. If we allow that 
the eponym chronicle ends with the death of Shamshi-Adad I, and the 
final 20 or more names in the karum-texts cannot be identified, the end 
of karum II will have been roughly in year 11 or 12 of Shamshi-Adad 
I. 13 This means that the restoration of a few more fragmentary entries 
in the Assyrian king-lists contributes significantiy to the second major 
source for the establishment of an Ancient Near Eastern chronology. 14 
Yet here too, it must be conceded that the later periods are better doc- 
umented and more reliable, than both the earlier era, and the neigh- 
bouring regions. These lists are the spine of all dates stretching back 
to the 12th century, complemented by eponym-lists (generally complete 
fort he first half of the first millennium), 13 and themselves anchored 
absolutely with a solar eclipse, dated to 15 June 763 BC, recorded in 
the eponym-lists. 16 The general compatibility of the versions of the 
Assyrian king-lists thus leaves only small margins in dating the reigns 
of the Assyrian kings back to the 12th century. 1 ' 

13 For details, cf. Veenhof, Akkadica 119-120, esp. 139 and Veenhof (n. 12), 61-62. 
Veenhof dates the accession of Shamshi-Adad to the year 1833 BC, on the basis of 
the Middle Chronology. Important is also his observation that the Distanzangaben of the 
later Assyrian tradition relate to Irishum I and Shamshi-Adad I, meaning that these 
are therefore probably realistic, effectively negating the hypothesis that the Distanzangaben 
are not chronologically relevant. Following the traditional Middle Chronology, Veenhof 
specifies the following Old Assyrian dates: the accession of Irishum I in 1974 BC and 
the death of Shamshi-Adad I in 1776 BC. 

14 On the king-lists, cf. above all, the article by A. K. Grayson, "Konigslisten und 
Chroniken. B. Akkadisch", RIA VI: 86-135; by comparison, the Sumerian lists are less 
useful, cf. D. O. Edzard, "Konigslisten und Chroniken. A. Sumerisch. 1.5. Historischer 
Wert", RIA VI: 81-82. 

15 Fundamental for the Neo-Assyrian eponyms is A. Millard, The Eponyms of the 
Assyrian Empire 910-612 BC (Helsinki, 1994); for the Middle Assyrian period, H. Frey- 
dank, Beitrage zur mittelassyrischen Chronologie und Geschichte (Berlin, 1991) is indispensable. 
For the more recent eponym lists, cf. the literature cited by Gasche et al., Dating, 47 
n. 204. 

16 The course of this total solar eclipse and other data is easily accessible on Espenak's 
NASA eclipse website: For the specifications of 
this solar eclipse, cf. also Stephenson, Eclipses, 126-127. 

17 For this, cf. J. Boese & G. Wilhelm, "Assur-ddn I., Ninurta-Apil-Ekur und die mittel- 
assyrische Chronologie", W^KM 71 (1979), 19ff; the error margin for Ashshur-resha-ishi 
I (1132-1115 BC) towards the end of the 12th century is +/- 2 years. 


In comparison with the Assyrian sources, the Babylonian material is 
far poorer. The Babylonians did indeed move relatively rapidly to a 
more elegant and relatively more accessible means of dating, namely 
simply counting the years sequentially according to the king's reign, 
which became the norm as early as the first half of the 14th century. 18 
However, there is a comparative dearth of written sources, 19 so that 
the Babylonian chronology must be established by synchronisms with 
Assyrian history, where works such as the synchronic history are impor- 
tant — if partial 20 — witnesses. 

The general situation is more or less comparable in all of the other 
areas, as a consistent sequential system of dates was not used anywhere, 
or is at least not preserved. This means that any attempt to establish 
a precise chronology for the history of the ancient Near East must rely 
initially on the Assyrian data, and only then on that of their neigh- 
bours, and their predecessors. Once these relative chronologies have 
been established, the second step is to identify the best possible syn- 
chronisms with the Assyrian chronology. Due to the multitude of sources, 
and especially the diplomatic correspondence of certain periods, it is 
possible to establish not merely one-to-one links, but actually to weave 
a fabric of relations which can actually be established without any inter- 
nal contradictions. 

Arriving at absolute dates depends upon a very different and diverse 
set of conditions. For the first millennium, people and events can be 
dated with near precision, as the uncertainty is highly constrained. For 
the earlier periods, some closed blocks can be isolated in which the 
relative chronology is certain, but anchoring these to the absolute 
dating of the first millennium leaves a considerable margin of error. 
There is a significant contrast with respect to the precise chronologi- 
cal synchronisations between Mesopotamia and Egypt for the first and 
second millennia. For the first millennium BC, where the Assyrian data 
is reliable and precise, the correlation difficulties lie with Egypt, whereas 
before this, the reverse is true. 

18 Cf. Brinkman (n. 3), 402f.; probably during the reign of Kadashman-Enlil II. 

19 A. K. Grayson, "Konigslisten und Chroniken. 3. Kings Lists", RIA VI: 89ff. offers 
a survey of the Babylonian sources; along with the fragments of a synchronic list, 12 Iff. 

20 This covers the period from the middle of the 2nd millennium through Adad- 
nerari III in the 8th century; edited by A. K. Grayson, Assyrian and Babylonian Chronicles 
(Locust Valley, 1975). 


Despite the millennia during which these civilisations existed, docu- 
ments confirming direct contacts and exchange between Mesopotamia 
and it neighbours on the one hand, and Egypt on the other, are rather 
rare and restricted to certain historical phases. Although one can assume 
that at all times there will have been people who were travelling back 
and forth or trading between these lands, this has left virtually no 
trace in the sources. We only have access to major state activities — 
diplomatic contacts or military engagements — which were generally 
exceptional. 21 The lucky find of the Amarna archive appears to be a 
period of particularly intensive contacts, but may actually have been 
repeated at very different times. The multi-facetted diplomatic archives 
from Hattusha can serve as a hint that cum grano sails the same conti- 
nuity of epistolary exchange may have taken place with other powers, 
and even in times which appear to us quite obscure. 22 

Despite such obstacles, there is a general consensus among scholars 
on the basic framework of the chronological framework which can be 
established using the basis provided by the Assyrian records, including 
also the reduction of ca. one decade in the Assyrian chronology pro- 
posed some time ago. 23 Another constraint is the revised model pro- 
posed by H. Gasche et al., which would result in a significant reduction 
even with respect to the Low chronology. Affecting the ancient Near 
East up to the middle of the second millennium BC, it touches a period 
for which the Assyrian sources do not provide complete cover. 24 

21 The seemingly comprehensive documentation in the Mari archives does not reveal 
one single direct link to Egypt, cf. A. Malamat, Mari and The Early Israelite Experience 
(Oxford, 1989), 6 If. c. n. 125. An informative survey on the geographical horizon of 
international relations in the Amorite period will be found in B. Lafont, "Relations 
intenationales, alliances et diplomatic au temps des royaumes amorrites", Amurru 2 
(2001), 213ff, where Egypt likewise fails to appear. 

22 The first fragment of a cuneiform letter, presumably from the diplomatic corre- 
spondence of Ramesses II, was found in 2003 in the excavations at Pi-Ramesses in 
the Delta, cf. E. B. Pusch & S. Jakob, "Der Zipfel des diplomatischen Archivs Ramses' 
II.", A&L 13 (2003), 145-153. 

23 Fundamental is the contribution by Boese & Wilhelm (n. 17) as well as the fur- 
ther and generally positive reception with the literature to be found by Freydank 
(n. 15), 11 n. 3. Freydank (n. 15; 34) confirms that although no definitive conclusions 
can be drawn from what is hitherto known from the Middle Assyrian eponyms, he 
tends to assign the two kings relatively short reigns. 

24 The arguments in Gasche et al., Dating, are supported not only by philological 
historical sources, but fundamentally based on archaeological criteria. For a critical 
evaluation, cf. G. Colbow, "Syrian Chronology in the Old and Early Middle Babylonian 
Periods", Akkadica 119-120 (2000), 103-116. 


For the period from the end of the 14th century, the dates of this 
new proposal do not differ significantly from those used hitherto, as a 
comparison between the Boese/Wilhelm shortened Assyrian chronol- 
ogy and the "ultra-low chronology" produced by Gasche et al. shows. 25 
The only point up for discussion between Tiglathpileser I (1114-1076 
BC) and Ashshur-nerari II (1414—1408 BC) is the reign of Ashshur- 
dan I, and his predecessor Ninurta-apil-Ekur, which has led to the pro- 
posed reduction of Assyrian chronology by 10 years, as mentioned. 26 
The interpretation of the expression tuppishu is less significant, since it 
concerns only the insertion of an additional year. 27 

The further back in Assyrian history one goes, the greater the diver- 
gences and the greater the room for differing chronological assump- 
tions so that Gasche et al. favour bringing the end of the First Dynasty 
of Babylon down to 1499, rather than 1531 or 1595 BC. At present, 
from the philological standpoint, only the statements of the periodic 
intervals (Distanzangaberi) can serve as control mechanisms. Unfortunately, 
there is no consensus on their value and chronological relevance. Gasche 
et al. logically argue that the temporal intervals preserved as Distanzangaberi 
are either incorrect or they try to provide a specific interpretation for 
the relevant interval. 28 An inscription of Shalmaneser I concerning his 

25 Compare the table in Boese & Wilhelm (n. 17), 38 with that of Gasche et al., 
g, 62. The work by Freydank (n. 15; 188-189), offers another survey of the reigns 

of the Middle Assyrian kings following the standard dates and the shortened Assyrian 
chronology. In the following, we follow the shortened chronology offered by Boese & 
Wilhelm, as this provides better data when compared with Egyptian chronology. 

26 For the details, cf Boese & Wilhelm (n. 17), 23ff; the shortening was not adopted 
by Gasche et al., Dating, 51 c. n. 223, probably because the collation of the relevant 
passage (cf. J. A. Brinkman, Or 42 [1973], 300 n. 15 and particularly on this, cf. Boese 
& Wilhelm (n. 17; 24) in the "Nassouhi-Kinglist" (= Copy A of King List No. 9, 
J. A. Brinkman, RIA VI: lOlff; the passage is in III 41f, RIA VI: 111) was interpreted 
as a confirmation although the reading — "46" — there, which Gasche et al., Dating, 51, 
used their table is described as "expressively optimistic", cited from Boese/Wilhelm 
and not the actual "26+ [x]" originally published by Brinkman, and cited by him again 
in the RIA VI: 111. 

2/ Cf. the references cited by J. A. Brinkman, RIA VI: 112, who translates "his 
tablet", for Ninurta-tukulti-Ashshur and Mutakkil-Nusku; W.v. Soden, AHw, 1304b, 
assumes that the corresponding entries are "short reigns of less than a year", while 
Freydank (n. 15), 33-34 assumes either a form of coregency or a distinction of the 
actual exercise of power during the formally still existing reign of Ashshur-dan I, 
whereby the corresponding entries in the king lists do not record an independent reigns. 
Gasche et al., Dating, 53-54, likewise reckon with years. 

28 Cf. Gasche et al., Dating, 57: "Our analysis of the Distanzangaben makes it appar- 
ent that no firm chronological conclusions can be drawn from them." This statement 
only applies to the model they propose. 


reconstruction of the Ekursagkurkurra specifies a period of 580 years 
separated his reign from that of Shamshi-Adad I. Understood in this 
fashion, this would be an obstacle to a major chronological change. 29 
The authors therefore propose that the passage should be interpreted 
as meaning that the number "580" does not refer to the interval between 
Shamshi-Adad I and the author Shalmaneser I, but also that the pre- 
ceding period of 159 years between the construction of the temple by 
Irishum I and its first renovation by Shamshi-Adad should be sub- 
tracted from the figure, 30 and therefore the 580 years thus represent 
the entire period between the construction and reconstruction. This 
necessarily leads to a substantial reduction in the chronology, a cen- 
tury and a half in fact. 31 

It must be stated that this significant reduction not only entails sug- 
gesting a reduction of the length of the Assyrian period of almost 200 
years, 32 whereby roughly a quarter of a century will have passed since 
the end of the Third Dynasty of Ur, 33 but also that a close link between 
the dynasty ruling in Yamkhad probably allows a link between the Old 
Hittite Period and the earlier periods, 34 and that this simply cannot be 
reduced at will. 35 The links in the chronological system of the first half 

29 Simply rejecting the Distanzangaben completely is apparently not appropriate as 
can be surmised from the analysis of the dates by Boese & Wilhelm (n. 17), 29ff.; 
indeed Veenhof 's demonstration, based on the recendy discovered Old Assyrian limu- 
Lists, is highly significant (cf. Veenhof n. 12; 139, c. n. 7). 

30 It is precisely this source, the reliability of which can be assumed; cf. the last 

31 For an analysis of the data in a diametrically opposed sense, cf. now Eder (n. 8), 

32 Veenhof (n. 12), 139f indicates that the period from Irishum I, year 1, to the 
death of Shamshi-Adad I was exactly 199 years, resulting from the recently discovered 
fozM-Lists allowing an overlap with the eponym-lists from Mari. 

33 Cf. Veenhof (n. 12), 141. 

34 Based on the genealogy beginning with Samsuiluna of Babylon, who was at least 
partially contemporary with Abba'el and over the immediate descendents Yarim-Lim 
II — Niqmepa — Yarim-Lim III — Hammurapi, who will have a contemporary of Hattushili 
I or Murshili I. 

35 In addition there is a solar eclipse listed in the Mari chronicle for the year after 
the birth of Shamshi-Adad I, for which C. Michel & P. Rocher, "La chronologie du 
IP millenaire revue a l'ombre d'une eclipse de soleil," JEOL 35-36 (1997-2000), esp. 
124 proposed that the most likely match would be the eclipse of 1795 BC. Taking 
the Distanzangabe separating Shalmaneser I and Shamshi-Adad I, it follows that 1234 
BC + 580 years = 1814 BC, placing the accession at the age of 18, resulting in a 
birth around 1832-33 which matches quite well with the solar eclipse of 24 June 1832 
BC. When revising her position to take account of the recent dedrochronological infor- 
mation, C. Michel ("Nouvelles donnees pour la chronologie", NA.B.U. 2002/1, 17-18) 


of the second millennium between Assyria and Babylon, with the death 
of the Assyrian king Shamshi-Adad I during the second decade of the 
reign of the Babylonian king Hammurabi 36 ultimately allows the syn- 
chronisation of Hittite history with Mesopotamian history through the 
campaign of Murshili I leading to the end of the First Dynasty of 
Babylon in year 31 of Samsuditana. Dating this event to the year 1499 
BC creates major problems for Hittite history. 37 

The question of the form of the Assyrian calendar is less serious than 
the differing evaluation of the Distanzangaben, but nevertheless impor- 
tant. The issue is whether it depended upon lunar months and whether 
or not intercalary months were used to match the solar year, 38 and 
thus whether or not calculations lead to a difference in the number of 
years in the chronology with respect to solar years. Fortunately, there 
is evidence favouring the use of solar years in Assyria during the sec- 
ond millennium and not just later. 39 

For the present purpose, these points do not have any significant 
impact since there are no direct synchronisms before the Amarna 

shifted her support to the identification of this eclipse. It is difficult to believe that the 
dates from such different sources could merely coincidentally converge into such a 
coherent chronology. It follows that a shortening of the chronology as proposed by 
Gasche et al., is excluded. Instead, this seems to suggest a chronological model slighdy 
higher than the "Low Chronology" and slighdy lower than the "Middle Chronology". 
Michel (above) proposes shortening the "Middle Chronology" by 15 years, and some 
such variant would be entirely compatible with the Hittite royal succession, and par- 
ticularly that of the Middle Hittite period. 

36 This applies although the precise year of the death of Shamshi-Adad I is still dis- 
cussed, as it may have been in year 12, 13 or 17 of the reign of Hammurabi; cf. 
Gasche et al., Dating, 64 and the literature cited in n. 265, as well as D. Charpin, in: 
D. Charpin, D. O. Edzard & M. Stol, Mesopotamia: Die Altbabylonische £eit (Fribourg: 
OBO 160/4, 2004), 155-156 (n. 713), 193. 

37 As is the new suggestion by Gasche et al., Dating, 64f, 77ff; they did not attempt 
to integrate the difficulties of Hittite history between Murshili I and Tutkhaliya I (cf. 
Gasche et al., Dating, note 92). G. Beckman, "Hittite Chronology", Akkadica 119-120 
(2000), 19ff. discussed precisely these issues at length and came to the conclusion "that 
the Anatolian evidence does not favour the suggested downward revision of the date 
of the fall of Babylon". 

38 The practice is known since the Ur III period, and letters from the Old Babylonian 
period record an intercalary month based on royal decrees, but the continuous usage 
of some rules cannot be documented, at least partially because of the lack of sources. 
In Assyria, the regular use of calibrating by using the lunar calendar is only known 
from the 1st millennium, cf. H. Hunger & E. Reiner, "A Scheme for Intercalary 
Months from Babylonia", WZKM 67 (1975), 21ff. and H. Hunger, "Kalender", RIA 
V: 297-303. 

39 Cf. the negative position taken by Veenhof (n. 12) 147, joining the arguments in 
Gasche et al., Dating. 


period. 40 We can thus use the dates for the Assyrian kings as proposed 
by Boese/Wilhelm as the basis for the presentation of the synchronisms 
with Egyptian history. 

The earliest contacts are recorded in the Amarna letters 41 which refer 
to earlier events, and specifically to the relations between Egypt and 
the Human Mittani state, the most important major power in Western 
Asia in the 15th century. After intense diplomatic contacts under 
Amenhotep II a dynastic marriage was arranged during the reigns of 
Thutmose IV and Artatama (I), which should be dated to the first 
decades of the 1 4th century according to the Egyptian chronology. The 
tradition was maintained by Shuttarna II and Amenhotep III who mar- 
ried a Mittani princess in his year 1 0, and later a daughter of Tushratta, 
who had since become ruler. In Shuppiluliuma I of Hatti, Tushratta 
met a dangerous foe in the first decade of the second half of the 1 4th 
century. In light of the complete absence of any original Hurrian or 
Mittani state sources, not much more chronological information can 
be gleaned from the information of the Amarna letters. 42 

It is only with the 14th century 43 letters of the Assyrian king Ashshur- 
Uballit I to Amenhotep IV (EA 15 and 16) that a real synchronism 
can be won from the Amarna letters. 44 There would appear to be con- 
siderable room for debate since the Assyrian ruled for more than 30 

40 The situation is different when considering the indirect relations which can be 
adduced from the interpretation of historical developments, as in, e.g., the question of 
whether the Hittite advance into northern Syria took place during a period of Mittani 
dominance or was in fact favoured by Egyptian advances; cf. J. Klinger, "Synchronismen 
in der Epoche vor Suppiluliuma I. — einige Bemerkungen zur Chronologie der mittel- 
hethitischen Geschichte", in: O. Carruba, CI. Mora & M. Giorgeri, eds., Atti del II. 
Congresso Intemazionale di Hittitologia (Pavia: Studia Mediterranea 9, 1995), 235ff. Ultimately, 
such reconstructions depend upon the chronology selected, and thus easily feed into 
circular logic. 

11 Cf. J. A. Knudtzon, Die El-Amarna Tafeln (Leipzig, 1915) or W. L. Moran, The 
Amarna Letters (Baltimore, 1992). 

42 For a summary, cf. Ktihne (n. 6), and St. de Martino, "II regno hurrita di Mittani: 
profilo storico-politico", in: La Parola del Passato LV (2000), 25ff. and with particular 
reference to the Amarna correspondence, B. M. Bryan, "The Egyptian Perspective on 
Mittani", in: R. Cohen & R. Westbrook, eds., Amarna Diplomacy. The Beginnings of 
International Relations (Baltimore, 2000), 71-84 and P. Artzi, "The Diplomatic Service 
in Action: The Mittani File", idem, 205-211. 

43 Assyria will have freed itself from Mittani hegemony shortly before the end of 
the 15th century, which allowed Ashshur-bel-nisheshu (1409—1400 BC) to renew the 
alliance with Babylon, under Kara-indash. 

44 Mentioning an exchange of letters between Ashshur-nadin-ahhe II and Amenhotep 


years, 4,1 but the contents of the first letter place it relatively near the 
start of his reign (1353 BC). 4b Thereafter follows a gap of several cen- 
turies in the historical record of direct contacts between the two powers. 

For further links between the cuneiform-using states and Egypt, we 
can turn to Babylonian and Hittite sources. As preserved, the Amarna 
correspondence documents contacts over several generations — as in let- 
ters between the Babylonian king Kadashman-Enlil I and Amenhotep 
III, and their successors Burnaburiash II and Amenhotep IV where the 
Egyptian must have come to the throne during the reign of the sec- 
ond Babylonian king, as the latter is in touch with both of these 
pharaohs. The preserved letters contain indications that of the earlier 
Kassite kings, Kara-indash had contact with Thutmose IV, and Kurigalzu 
I with Amenhotep III. 47 Since the reigns of the Kassite Kings before 
Kadashman-Enlil are unknown these vague hints appear unpromising, 
yet one can actually create a chronological framework since we know 
that Kara-indash reigned at the same time as the Assyrian king Ashshur- 
bel-nishu (1407-1399 BC), which puts some limits for the not partic- 
ularly long reign of Thutmose IV. 48 

The sources are more precise for Kadashman-Enlil I 49 and Burna- 
buriash II, 50 as the reign of the latter is known, and the synchronic 

45 In the absence of detailed information about the temporal relationship between 
the predecessors of Eriba-Adad I and Amenhotep III and his successor Amenhotep 
IV, we can only deduce that the exchange must have taken place around 1381 BC. 

46 Beckerath, Chronologie, 61 places this "shortly after his accession to the throne", 
which he dates to 1355 BC, and thus according to his chronology, dating the acces- 
sion of Amenhotep IV to the end of 1 35 1 BC, allowing an indirect synchronism, as 
EA 15 does not name the pharaoh. Quite apart from this, there is sufficient time in 
the reign of Amenhotep IV for the second letter to have been addressed to him, despite 
the uncertainties with respect to the reading of the name. 

47 Above all, Burnaburiash II to Amenhotep IV in EA 10, 8ff. The details of the 
Babylonian succession add to the uncertainty, but do not have any significant impact 
in this context, cf. Brinkman (n. 3), 169 and J. A. Brinkman, "Istanbul A. 1998, Middle 
Babylonian Chronology, and the Statistics of the Nippur Archives", ^A 73 (1983), 
67-74. For the beginnings of the correspondence with Babylon, cf. C. Kiihne, Die 
Chronologie der internationalen Korrespondenz von El- Amarna (Neukirchen-Vlyn: AOAT 17, 
1973), 52f. n. 244. 

48 Recently, J. A. Brinkman, "Nazi-Marruttas", RIA 9: 190b, placed Kadashman- 
Harbe I ca. 1405, and Kurigalzu I ca. 1390 BC; cf. also the chronological table by 
Sassmannshausen (n. 4), esp. 67-69. 

49 The assumption of a 15-year reign is not, however, compelling, as Brinkman 
(n. 3), 142f. c. n. 27 specifically stresses. 

50 Following Brinkman, this would be 1359-33 BC, as recently in RIA 9: 190b, i.e., 
the same dates as those in his table published in Oppenheim & Reiner (n. 2), 338. It 
should be noted that Brinkman (n. 3), 32 n. 89 already observed the proposed dates 


history links the end of his reign with the Assyrian chronology. 51 As he 
actually corresponded with Amenhotep III,' 2 but the exchange of let- 
ters only began after his 30th year, 53 the entire reign of Amenhotep 
IV would thus parallel that of Burnaburiash II, who must have reigned 
until the accession to the throne of the successor of this pharaoh, and 
thus to the year 1323 BC according to the shortened chronology. 54 His 
accession to office must then have taken place in 1349 BC, when 
Amenhotep III would still have been reigning according to the syn- 
chronism of the Amarna letters. However, some of the numbers remain 
uncertain due to factors which have not yet been eliminated in the 
reigns of the Kassite rulers, and thus the figures cannot be taken as 
absolute limits. Discrepancies with Egyptian dates 55 can easily be explained 
as lying in the Babylonian sources. We must therefore confirm that the 
known synchronisms between Egypt and Babylon cannot provide and 
exact chronological fix points due to the imprecision inherent in the 
frame of reference, or in the uncertainties in the chronological frame- 
work itself, although they do place some further limits on the range of 
variation. 56 

According to sources in internal Hittite history, contacts between the 
Egyptian and Hittite kings were so intense that a treaty already clarified 
relations between the two in the 15th century. Fragments of this accord — 
known as the Kurushtama-treaty — are preserved, 5 ' and its significance 

ultimately depend upon the reigns of the Assyrian kings, and that aside from a few 
uncertain factors mentioned there that any reduction must be accompanied by a cor- 
responding change there; however, there still remains a margin of +/— 5 years. 

" For the details cf. Brinkman (n. 3), 418ff. 

' 2 This is the most probable interpretation of EA 6 according to the collation by 
Kiihne (n. 47), 129 c. n. 642. 

53 Cf. most recently, Beckerath, Chronologie, 66. 

>4 Cf. also Brinkman (n. 3), 6 n. 1. 

55 The reduction in the Babylonian chronology in this case would stand in contra- 
diction to the assumed accession of Amenhotep IV ca. 1351/50 BC, as Burnaburiash 
would only have ascended to the throne in the following year, and thus a synchro- 
nism with the older of the two pharaohs would be excluded, although precisely this 
is fact reliably attested. The margin of uncertainty in this phase remains the +/— 5 
years mentioned. 

>( ' A fragmentary economic text found during the excavations at Assur in 2001 (Ass. 
200 l.D— 2217) documents a direct synchronism between Melishipak of Babylon (ca. 
1186-1172 BC following J. A. Brinkman, "Meli-shipak", RIA VIII: 52) and Ninurta- 
apil-Ekur of Assyria (for whom, cf. E. Cancik-Kirschbaum, AoF 26 [1999], 215f£); cf. 
the preliminary report by E. Frahm, MDOG 134 (2002), 75. 

57 The few sources available are assembled under CTH 134. KUB 40.28 (= 134.C) 
is a fragmentary text preserved in a Middle Hittite copy. A. UnaPs (RIA VI: 373) 


recorded later in the context of the Deeds of Shuppiluliuma I, 58 yet 
the hints do not suffice to provide a precise context. 59 According to the 
sources, the only possible Hittite ruler would be Tutkhaliya I, 60 as oth- 
erwise the Hittite history of this period was dominated by internal 
conflict with a correspondingly weak foreign policy, and thus actions 
in Syria cannot be expected. 61 The pharaoh who was party to this 
treaty can only be traced through the Hittite sources, 62 and for the 
moment these are inadequate to define the period when Tutkhaliya I 
reigned, 65 beyond specifying that it might have been around the third 

speculation that these Kurushtama-fragments could merely be part of the Egyptian cor- 
respondence of Shuppiluliuma I is thus impossible. 

58 DS Frag. 28 (cf. H. G. Giiterbock, "The Deeds of Shuppiluliuma as Told by his 
Son Murshili II", JCS 10 (1956), 41-68, 75-98, 107ff.). A great many details are still 
unresolved, and thus it is impossible to state with certainty that it merely a question 
of a single treaty and not possibly a series of international accords; for this issue, cf. 
D. Stirenhagen, Paritatische Staatsvertrage aus hethitischer Sicht (Pavia: Studia Mediterranea, 
1985), 22ff. Equally uncertain is the exact placement of those fragments 29 and 30 
discussed by H. G. Giiterbock of DS, which both mention Egypt. 

59 Likewise belonging to the Middle Hittite period in the draught letter in Hittite, 
where neither the name of the Hittite king nor that of the Pharaoh to whom it was 
addressed is preserved. Given its date, the text can be assigned to the period before 
Shuppiluliuma I. As preserved, it can be deduced that this letter followed another 
exchange and the Hittite kings complains of a deterioration in relations which has now 
been resolved. Tutkhaliya I is one possible author, but this remains pure speculation 
in the absence of further information. For the text itself, cf. E. Edel, "Bo 92/129, ein 
neues Brieffragment in hethitischer Sprache der Korrespondenz zwischen Agypten und 
Hard", ZA 86 (1996), 114-7; since published as KBo 31.40. 

60 There is still some doubt about the number of kings named Tutkhaliya who 
reigned in the period immediately before Shuppiluliuma I. In this case I assume the 
sequence: Tutkhalija I — Arnuwanda I — Tutkhalija II — Shuppiluliuma I; cf. also simi- 
larly, H. Klengel, Geschichte des Hethitischen Reiches (Leiden: HdO I. 34, 1999), 103. 

61 Cf. the literature cited by Klengel (n. 60), 106f. (on source [A8]) c.n. 93f, who 
likewise stresses this identification is purely hypothetical. 

62 W. Helck, Die Beziehungen Agyptens zu Vorderasien im 3. und 2. Jahrtausend v. Chr. 
(Wiesbaden: AA 5, 1971 2 ), 166 suggested either Thutmose III or Amenhotep II, but 
did not exclude Thutmose I. In any case, it should be evident that the events detailed 
here are incompatible with the proposed date of 1499 BC for the campaign of Murshili 
I to Babylon, as they took place more than a century later, and thus well into the 
14th century. The synchronism between Idrimi and Murshili I recently proposed by 
W. van Soldt, "Syrian Chronology in the Old and Early Middle Babylonian Period", 
Akkadica 119-120 (2000), 111 or between Idrimi and Zidanta I, by Kuhne (n. 6), 214 
n. 67 are incompatible with the chronological scheme proposed here, as is the attempt 
to make a historical link between Murshili I and Thutmose I; for this, cf. F. Zeeb, 
"The History of Alalah as a Testcase for an Ultrashort Chronology of the Mid-2nd 
Millennium BCE", in: Hunger & Pruzsinszky (n. 4), 9 If. 

63 The proposed reigns of some 20-25 years each for the three kings before 
Shuppiluliuma I in the chronological table of the exhibition catalogue, Die Hethiter und 
ihr Reich (Stuttgart, 2002), 312ff. are largely hypothetical or rely upon an historical 


quarter of the 15th century. 64 We can, however, define the number of 
generations separating Tutkhaliya I and Shuppiluliuma I, since the lat- 
ter was the son of the former's grandson (Tutkhaliya II). 65 Given the 
difficulties of counting generations, the internal between them should 
be more than 50 and less than 100 years. 66 

Similarly difficult is precisely delimiting the reign of Shuppiluliuma 
I, 6 ' whose long reign will have stretched from that of Amenhotep III 
until perhaps the reign of Aya. 68 It is only certain that he was in direct 
contact with Amenhotep IV. Whether he was also in touch with his 
father depends upon identifying the pharaoh concealed behind the 
apparently unreliable cuneiform transcription Hurija in EA 41 written 
by Shuppiluliuma I. 69 Any further attempts to establish a more exact 

image dependent upon the short chronology, which simply cannot be assumed given 
the very limited historical information at our disposal; this also necessarily applies to 
the proposed accession date ca. 1420 BC for Tutkhalija I. 

64 Dating Thutmose I to the first decade of the 1 5th century means that he should 
be most probably understood as a contemporary of Tutkhaliya I, as the now reliable 
reconstruction of the royal succession in the Middle Hittite period tends to bring 
Tutkhaliya I closer to Shuppiluliuma I than was the case earlier. On the other hand, 
the land donation documents definitely attest that allowance must be made for more 
generations between the reign of Tutkhaliya I and the end of the Old Hittite period 
(or the beginning of the Middle Hittite period, with the reign of Telpinu as the thresh- 
old) than had been appreciated earlier. 

65 A detailed discussion of the Middle Hittite king lists is impossible here for obvious 

66 For a detailed account of the specific problem as concerns Hittite history, cf. 
G. Wilhelm, "Generation Count in Hittite Chronology", in: Hunger & Pruzsinszky 
(n. 4), 71-79; Eder (n. 8), 224ff. likewise reveals the great differences possible in the 
interpretation of the material. 

67 Cf. Klengel (n. 60), 147: "The precise moment of the accession cannot be fixed;" 
the various possibilities were discussed by G. Wilhelm & J. Boese, "Absolute Chronologie 
und die hethitische Geschichte des 15. und 14. Jahrhunderts v. Chr.", in: Astrom, Acts, 
73-117, esp. 76ff. themselves pleading for a much later accession, possibly even after 
the accession of Amenhotep IV, whose reign they set at 1343-1322 or 1318 BC. 

68 Most recently discussed in detail by Wilhelm & Boese (n. 67) suggesting a significant 
reduction in the length of the reign rather than the 40 years traditionally accepted. 
This has not been accepted as the sources simply do not allow for a definitive con- 
clusion, but the issue must remain open. Cf. also G. Wilhelm, "Probleme der hethiti- 
schen Chronologie", OL£ 86 (1991), cols. 47 Iff. The issue was most recently discussed 
by J. Freu, "La chronologie du regne de Suppiluliuma: Essai de mise au point," in: 
P. Taracha, ed., Siba Anatolica (Warsaw, 2002 = Fs Popko), 87ff. He opposes a dras- 
tic reduction in the reign of Shuppiluliuma I, and assumes a 30-40 year reign — 
dependent upon a coregency between Amenhotep III and Amenhotep IV (esp. 107). 

69 A decision depends largely upon the author's historical interpretation, but there 
is a tendency to exclude Amenhotep III, probably following Wilhelm & Boese (n. 67), 
96ff.: W. L. Moran (n. 41), 114f and Klengel (n. 60), 139 (on source [A 10]). 


chronological ordering of the historic events thus depends upon the 
interpretation of the Dakhamunza-affair, that is of who made the aston- 
ishing offer of a dynastic marriage to the unbelieving Hittite king — 
and a generally accepted solution has still not emerged. 70 The widow's 
name is not preserved in the Hittite version, and the throne name, 
Nibhurunia, of the dead pharaoh is recorded in the cuneiform version 
of the Deeds of Shuppiluliuma I, but this was only written down on 
the orders of his son Murshili II. The question is therefore identifying 
the name of the pharaoh whose widow sent the proposal to Shuppiluliuma 
I. 71 Can the cuneiform Nib at the beginning of the name be only an 
Egyptian nb, or is a nf(r) also possible? 72 According to my judgement, 
in the absence of a sufficient quantity of cuneiform attestations which 
could offer a corresponding orthographic principle to which one could 
refer, the linguistic and orthographic arguments simply cannot be 
resolved. 73 Quite apart from this, there is no means of knowing whether 
a Hittite copyist faced with an unfamiliar name written with what was 
at that time the quite alien sign /nap/ may not have chosen to divide 
the signs syllabically as NI-IB. Given that the quite adequate tran- 
scription of Nb-hprw-R c as m Ni-ib-hu-ru-ri-ja (KUB 34.24+ rev. IV 
18) 74 is in principle correct, we can assume that the throne name of 

70 The literature here is almost endless, we cannot go into detail here, as for some 
time now the Hittite sources have simply not shed any new light on the matter. The 
most recent discussion of the question is that of M. Gabolde, D'Akhenaton a Toutdnkhamon 
(Lyon, 1998), to which we refer here. M. Gabolde offers a summary of his thesis: "Das 
Ende der Amarnazeit", in: A. Grimm & S. Schoske, eds., Das Geheimnis des goldenen 
Sarges (Munich, 2001), 9—42. Basing himself on new epigraphic finds and new inter- 
pretations of previously neglected (or differently interpreted) materials, he identifies the 
widow as Meritaten who replaced Nefertiti near the end of the reign of Amenhotep 
IV, and when he died, turned to Shuppiluliuma I. This compels him to identify 
Nibhururii as Nfr-hprw-R' Amenhotep IV, and to assign the letter EA 9 to him. For 
this, cf. M. Eaton-Krauss & R. Krauss, [review of Gabolde], BiOr 58 (2001), col. 96 
and furthermore W.J. Murnane, OLZ 96 (2001), 1 Iff. 

" All of the preserved texts relevant to the remarkable episode were assembled by 
Th.P.J. van den Hout, "Der Falke und das Kucken", Z A 84 (1994), 60-88, esp. 61f. 

72 For a long time, a reading of nb- favoured Tut'ankhamun, but this changed when 
Krauss, Amarnazeit, 9ff. argued in favour of nfr-, and thus for Amenhotep IV. 

''■' Definitely favouring the first possibility is most recently T. Bryce, "The Death of 
Niphururiya and its Aftermath", JEA 76 (1990), 97-105, whereby he shares what can 
be viewed as the Hittitological consensus, as argued by Wilhelm & Boese (n. 67), 100fi; 
cf. also van den Hout (n. 71), 84f; E. Edel, Die dgyptisch-hethitische Korrespondenz aus 
Boghazkoi in babylonischer und hethitischer Sprache //(Opladen, 1994), 23 n. 3; and G. Meyer, 
GM 126 (1992), 87-92. 

' 4 In contrast KFSo 5.6 obv. Ill 7 has m Pi-ip-hu-ru-ri-ja-a; an error of Pi for NE? 


Tut'ankhamun was intended, as alternative readings are difficult to jus- 
tify, and actually rely upon a further series of assumptions. Assuming 
this identification of the pharaoh, we can project a series of dates for 
the end of the reign of Shuppiluliuma I. After the victorious campaign 
with which the Hittites responded to the murder of the Hittite prince 
Zannanza, 7 ' 1 Egyptian prisoners brought the plague to Hattusha, and 
Shuppiluliuma I himself died shortly thereafter, possibly due to the epi- 
demic. 76 The death of Shuppiluliuma I — and the accession to the throne 
of his two sons Arnuwanda II and Murshili II — can thus be tied to 
the Egyptian chronology, and Shuppiluliuma I can only have lived for 
2—4 years after the death of Tut'ankhamun. On the other hand, this 
also enables us to verify the veracity of the interpretations due to the 
solar eclipse dating to the reign of Murshili II, whose reign can thus 
be given absolute dates. This means that the two systems must be 
formed in a fashion which allows an accession to the throne for Murshili 
II in 1321, 77 since the death of Shuppiluliuma I must be limited to 
either 1323-1322 or 1326-1325 BC. 7!i In contrast to the first men- 
tioned contacts between Egypt and the Hittites, the Dakhamunzu episode 
might — with all due caution — provide a date relevant to the entire 
chronological discussion, allowing a link for the two independent chrono- 
logical systems. 

For the decades after the Amarna period, the sources on Hittite - 
Egyptian relations vary widely. Although the age of Murshili II is among 
the most productive in terms of overall Hittite historiography, there 
would appear to be few real opening for the study of relations with 
Egypt. In terms of their political calculations, the Hittites did not feel 

73 The thesis argued by M. Gabolde, that Zannanza actually briefly reigned under 
the throne name of Smenkhkare is not probable; cf. the sceptical judgement of Eaton- 
Krauss & Krauss, BiOr 58 (2001), col. 96. 

"' As noted, what can be won from the Hittite sources was assembled by van den 
Hout (n. 71), 85ff. 

11 Cf. the detailed discussion by Wilhelm & Boese (n. 67), 105ff. Nevertheless, the 
assumption that the solar eclipse took place in the spring (ibid, pp. 106f.) is not oblig- 
atory as it actually relies upon what is actually a rather dubious restoration of KUB 
14.4 IV 25f. by E. Forrer, but cf. A. Gotze, K1F 1, 1930, 405. Therefore, nothing 
stands in the way of the total eclipse of 24 June 1312 BC. 

78 Favouring the latter date is van den Hout (n. 71), 88, who arrives at 17 years 
from the death of Nibkhururija = Tut'ankhamun to the solar omen (with 6 years of 
the reign of Shuppiluliumas I + 1 year of Arnuwanda II + 1 years of the reign of 
Murshili II), and thus assumes a death in 1325 BC; his calculations are, however, 
based on the solar eclipse of 13 April 1308 BC. 


either challenged or constrained by Egypt. This only changed significantly 
when Ramesses II came to the Egyptian throne, changing the politi- 
cal constellation during the reign of Muwatalli II, the son and succes- 
sor of Murshili II. /9 The temporary removal of the Hittite capital from 
Hattusha to Tarkhuntashsha had significant repercussions for the tex- 
tual record such that the reign of Muwatalli is badly documented. The 
lack of relevant documents can be interpreted as meaning that some 
of the documentation was not returned when Hattusha became the 
capital again under Muwatalli's successor. This may be the reason why 
that period of increasing contact which culminated in the battle of 
Kadesh, so heavily stressed in the Egyptian tradition, appears to be 
marginal in the Hittite tradition. A further consequence of our limited 
understanding of this period in Hittite history is that at the pinnacle 
of the Hittite -Egyptian conflict, the battle of Kadesh in year 5 of 
Ramesses II, we can only specify that his opponent was Muwatalli II. 
The treaty signed in the wake of the conflict in year 2 1 of Ramesses 
II was negotiated and sealed by Hattushili III. 80 Nor are there any 
sources from the Hittite capital Hattusha which can allow this decisive 
event to be dated, despite the abundance of letters from both king- 
doms dating to before and after the signing of the treaty, and despite 
the wealth of details we have for the internal history of the Hittites 
during this important age. 

Nevertheless, the Egyptian sources do allow some key dates, throw- 
ing light on the transfer of power from Urkhi-Teshshup (= Murshili 
III), whose reign was ended by the usurpation of Hattushili III during 
these 1 6 years. It is unclear how long after year 5 of Ramesses II, that 

79 The end of the reign of Murshili II can only be approximately delimited since 
for the later years of the reign the preservation of the annals is significantly worse than 
for the phase up to ca. regnal year 20. Nevertheless, we can be relatively certain that 
Murshili remained on the throne for some three decades. A. Unal, "Muwatalli II", 
RIA VIII: 524-527, esp. 524 assigns Muwatalli II a reign of "ca. 1290-1265 BC", but 
postulates that he either fell (or was mortally wounded) at the battle of Kadesh, and 
therefore dying then or shordy thereafter, yet he likewise dates this to ca. 1274 BC 
(ibid., 527) — which would be nearly a decade earlier. The source used by A. Unal in 
this connection, linking the death of the Hittite king with this campaign (citing ABoT 
57 obv. 8 is probably an error, as the passage, and as KBo 4.10+ obv. 40'ff. clearly 
confirms, relates to the removal of the capital and not to the death of the king), refer 
only to the death without placing this in any identifiable context. 

80 For the text of the treaty, cf. E. Edel, Der Vertrag zwischen Ramses II. von Agypten 
una! Hattushili III von Hatti (Berlin: WVDOG 95, 1997) prefaced (p. 1) with a short 
account of the prehistory of the treaty. 


Urkhi-Teshshup, the direct and legitimate heir of Muwatalli II, came 
to the throne as Murshili III, and how long he reigned. 81 It was dur- 
ing his reign that the removal of the capital was cancelled, for reasons 
beyond our knowledge. The sources are silent about Murshili's other 
activities, 82 and direct synchronisms with other kings cannot be proved. 83 
Nevertheless, a draught letter from the reign of Hattushili III to the 
Assyrian king — which could be either Adad-nerari I or Shalmaneser 
I 84 — mentions contacts between his predecessor and the Assyrian recip- 
ient of the royal letter, and the correspondence between Hattushili III 
and Ramesses II suggests that Murshili III was also in touch with this 
Egyptian king, although these cannot be documented either. Just how 
dependent the interpretations of the fragmentary letters are upon the 
assumed historical background is illustrated in exemplary fashion by 
the fragmentary draught of a letter (KUB 23.102). It is generally assumed 
that the intended recipient was Adad-nerari I, given the clear hints at 
a very tense relationship between Assur and Hatti, and the Hittite king 
dismisses any equality between the two. This suggests that the letter 
should be dated to a period of increasing Assyrian power, but a concrete 

81 The immediate circumstances of the usurpation and its possibly not entirely unan- 
imous acceptance have been frequently discussed, but the partial character of the 
sources available mean that the details still remain obscure. In view of the partial 
descriptions by one of the actors, namely the usurper Hattushili III, it is significant 
that there is a seal impression naming Muwatalli and the later king Murshili with his 
given name, which could potentially imply a relatively early attempt at securing the 
succession; for a detailed survey of the entire situation, cf. most recently, Ph. Houwink 
ten Cate, "Urhi-Tessub revisited", BiOr 51 (1994), 239ff. 

82 Summarized by Klengel (n. 60), 226ff. 

83 The reason for this may be that the lack of information forthcoming from the 
fragmentary sources available, e.g., the diplomatic correspondence cannot generally be 
assigned to or related to Murshili III. On the general tenor of the later sources on 
Murshili III, cf. also the literature cited Klengel (n. 60), 257 n. 501. 

84 This is the letter KBo 1.14; cf. A. Hagenbuchner, Die Korrespondenz der Hethiter. 2. 
Teil: Die Briefe mit Transkription, Ubersetzung und Kommentar (Heidelberg: THeth 1 6, 1 989), 
No. 195 (obv. 15f.) and A. Harrak, Assyria and Hanigalbat (Hildesheim: TSO 4, 1987), 
69ff, who pleads for Adad-nerari I while A. Unal, Hattushili III., Teil I, Hattushili bis 
zu seiner Thronbesteigung, Bd 1: Historischer Abrifi (Heidelberg: THeth 3, 1974), 6, favours 
Shalmaneser I. By contrast, following M. B. Rowton, "The Background of the Treaty 
between Ramesses II and Hattushili III", JCS 13 (1959), 1-11, or M. B. Rowton, "The 
Material from Western Asia and the Chronology of the Nineteenth Dynasty", JNES 
25 (1966), 249-258, the only possible candidate would be Adad-nerari I. This proposal 
is, however, based exclusively on an interpretation of the material based upon the con- 
tent, and a corresponding reconstruction of the historical context — which is by means 
the only possibility. The reference to Murshili III while avoiding the throne name 
Urhi-Tessup only permits us to conclude that one of his successors wrote the letter. 


synchronism does not follow. It is not irrelevant that we can hardly 
expect to hnd draught letters from the reign of Muwatalli II in the 
archives of Hattusha, 8 ' 1 as this favours attributing the letter to Murshili 
III. An uncertain but possible synchronism of the latter with Adad- 
nerari I would follow, leading to a throne change before 1264 BC (the 
death of Adad-nerari I). Were one to date the letter to Hattushili III, 
the length of the reign of Murshili would shrink further. 86 It is certain 
that Murshili III came to the throne after 1275, and remained on it 
until perhaps 1263 BC, but a shorter reign is more probable. The (at 
least) "7 years" during which Hattushili III allegedly remained loyal 8 ' 
according to his "Apology" need not be taken as a real measure of 
time; although possible, it cannot be independently verified. This means 
that the accession to the throne of Hattushili III cannot be determined 
with precision. Neither the Egyptian chronology nor Egyptian sources, 
including the correspondence of Ramesses II, offer additional aid. 88 As 
the synchronism with Shalmaneser I can only offer a date after 1263, 
then the letter KBo 1.14 becomes interesting. If this letter is from 
Hattushili III and addressed to Adad-nerari I, then — and only then — 
would we be able to place the beginning of the reign of Hattushili III 
in the period after 1268 and before 1264/3 BC. 89 

Although generally viewed as a letter from Tutkhaliya IV to Tukulti- 
Ninurta I, which would provide another hook for the end of the reign 

35 Following Hagenbuchner (n. 84), 263; for a discussion of the correct placing of 
the text, cf. Klengel (n. 60), 204 c. n. 304. 

86 A. Hagenbuchner, THeth 16, reads in obv. 16 [A-BI A-BI-J]A-ma ! A-BI-JA-ja 
"my father and [m]y [grandfather], however", yet the position of the -ma would be 
unusual, aside from the fact that a collation with the photo suggests that the sign 
should rather be read as BA, so that the point of departure should be A-BA A-BI-JA- 
ja, and thus exactiy that expression used in KBo 6.28 obv. 16 by Hattushili III with 
reference to his grandfather Shuppiluliuma I, during whose reign, Assyria first attempted 
to free itself from Mittani hegemony. By contrast, in his 9th regnal years his succes- 
sor Murshili II was forced to ward off an assault by Adad-nerari I which reached as 
far as the eastern bank of the Euphrates at Carchemish; the Assyrians managed to 
take Taide, Wasashatta's capital, as mentioned in KUB 23.102 obv. I 1. Regardless, 
Urhi-Tessup was the grandson of Murshili II. 

8/ Cf. H. Otten, Die Apologie Hattusilis III. Das Bild der Uberlieferung (Mainz: StBoT 
24, 1981), 20f: "And given the estimation of my brother, I refrained from any actions, 
and was obedient for 7 years" (rev. Ill 62; on the verbal form in question, cf. Most 
recently J. Tischler, HEG III, 421). 

88 The peace treaty with Hattushili III was sealed in regnal year 21 (i.e., 21/XI/1259 
BC) of Ramesses II, and was followed by a widely publicised dynastic marriage in reg- 
nal year 34. 

89 Cf. Boese & Wilhelm (n. 67), 36 n. 65. 


of Hattushili III, the fragmentary letter KUB 26.70 cannot really shed 
light on the matter. We must bear in mind that in terms of content, 
this would suggest an exchange of letters between Urkhi-Teshshup/ 
Murshili III and Shalmaneser I, but this can nevertheless have taken 
place after the usurpation by Hattushili III. 90 Even using the fragments 
of the diplomatic correspondence with Assur, we simply lack the nec- 
essary cornerstones required to date the reigns of the Hittite kings of 
the 13th century. 91 Hattushili III and Tutkhaliya IV reigned during the 
times of Shalmaneser I, but we cannot tell whether his 30 year reign 
began before the fall of Murshili III, or only thereafter, although the 
latter possibility is more probable. Tutkhaliya IV definitely reigned in 
Hatti when Tukulti-Ninurta I came to the throne in Assur. Whether 
he was still in office when Shuppiluliuma II became king cannot be 
confirmed as nothing can delimit the length of the reign of the last 
king known from the archives of Hattusha, according to which the 
Hittite empire ended sometime shortly after 1200 BC. 92 Nor are there 
any hints at contacts to any of the rapidly disappearing kings during 
the uneasy period after the murder of Tukulti-Ninurta, i.e., Ashshur- 
nadin-apli to Ninurta-apil-ekur and Ashshur-dan II. 

The correspondence with the Babylonian kings is ever worse pre- 
served than that with the Assyrians, who only allow a few additional 
points in a more precise temporal organization of the Hittite kings of 

90 For this, cf. primarily, S. Heinhold-Krahmer, "Zu Salmanassars I. Eroberungen 
im Hurritergebiet", AJO 35 (1988), 94 n. 181, and now Beckerath, Chronologic JVR, 26 
c. n. 144 (following up on Hornung's earlier work). 

91 Litde of value can be gained from the fragments (KBo 28.59—63) of a Middle 
Assyrian royal letter sent to Hattusha, but one can still appreciate that the subject is 
the behaviour of Sagarakti-Surias, of Babylon which suggests that one can assume that 
it was composed between the later part of the reign of Shalmaneser I and the begin- 
ning of that of Tukulti-Ninurta I; for such a reading, cf. W. von Soden, "Weitere mit- 
telassyrische Briefbruchstucke aus Hattusas", in: E. Neu & Ch. Ruster, eds., Documentum 
Asiae minoris antiquae (Mainz, Fs Otten, 1988), 346. The reference to Tutkhaliya does 
suggest dating the correspondence to his reign, yet this does not allow its chronolog- 
ical limits to be defined any further. Finally, it is unclear just what the reference to a 
period of "100 years" signifies (KBo 28.61 rev. 9'). 

92 The date from Emar for year 2 of the Babylonian king Melishipak — which cor- 
responds to the year 1187 BC, cannot be used for the abandonment of the Hittite 
capital Hattusha as it is by no means certain that the "Sea Peoples" were responsible 
for the end of Emar, cf. now Klengel (n. 60), 318 c. n. 35. The destruction of Ugarit 
thus remains the terminus ante quern, but this only allows an extension into the reign 
of Siptah, i.e., until 1197 BC (cf. the literature cited by Klengel (n. 60), 318, no. 33, 
as well as Ras Shamra-Ougarit XI, [1995] with the text published by S. Lackenbacher, 
RSO XI, 77ff.). 


the 13th century, allowing for some synchronisms supported by the 
sources. There is a letter from Kadashman-Turgu I to Hattushili III 
(KUB 3.7 1), 93 and these two would appear to have signed a treaty with 
each other, as indicated by a draught letter prepared at the Hittite 
court (KBo 1.10+9), but apparentiy during the reign of Kadashman- 
Enlil II and to be assigned to Hattushili III who mentions the acces- 
sion of the young Babylonian ruler in the text. 94 The change of power 
in the Kassite royal house is set at about 1264/63 BC, but from the 
Kadesh treaty it is clear that Hattushili III came to the throne before 
the year 2 1 of the reign of Ramesses II. Reducing this date by a decade, 
following the Assyrian regnal years, would appear to be difficult, but 
there nevertheless remains the margin of uncertainty, amounting to 
almost a decade for the Babylonian kings, to which repeated reference 
has been made, and which cannot be further reduced for the moment. 

93 For this, the best discussion remains E. Edel, Agyptische Arzte und agyptische Medizin 
am hethitischen Konigshof. Meue Funde von Keilschriftbriefen Ramses' II aus Bogazkoy (Opladen, 
1976), 123ff. 

94 Above we mentioned the indirect references relating to Sagarakti-Surias, from the 
later correspondence with Tukulti-ninurta I. Whether this Babylonian king's reign was 
contemporary with that of Tutkhaliya IV (the presumed recipient of the letter) is uncer- 
tain, as the relationship between the time the letter was written and the references to 
Tukulti-ninurta I who conquered Babylon during the reign of Kashtiliash IV (1232—1225 
BC) is not clear. It depends upon whether the events mentioned in the text took place 
at that time, or whether the letter mentioned events which had taken place at some 
point in the past. 



Stmt W. Manning 

1. Introduction: History of Field 

In the beginning, the historical chronology of Egypt was held to offer 
a test for the utility of the radiocarbon dating method; measurements 
were thus run on several ancient Egyptian samples and the ability to 
achieve ages relatively close to the historical age demonstrated that 
radiocarbon dating worked in approximate terms (or was not 'beyond 
reasonable credence'). 1 Egyptian samples thus comprised part of the 
original 'curve of knowns' published in Arnold and Libby 2 to show that 
the radiocarbon method worked, approximately, over the last several 
thousand years. 

Over the next few decades a number of radiocarbon ages were 
obtained on Egyptian samples. Egyptian chronology continued to be 
considered as the known age, and radiocarbon was being compared — 
tested. Radiocarbon technology through the 1960s was not capable of 
delivering ages of sufficient accuracy or precision to be of actual util- 
ity to Egyptologists. 3 In 1970 Save-Soderbergh and Olsson published a 
well thought out critical analysis of radiocarbon dates from Egypt. 4 
They highlighted problems of poor association between samples and 
presumed historical context (or age), of contamination, and of the need 

* The final text of this paper was submitted 19 August 2003. The paper employs 
the then standard IntCal98 radiocarbon calibration dataset. A new IntCal04 dataset has 
since been published in early 2005 [Radiocarbon vol. 46(3), 2004). Use of the new dataset 
would make only small and fairly insignificant changes to the figures and discussions 
in this paper. For a comparison of the two calibration curves for the period 500-3500 
BC, see Figure III. 1.6 at the end of this paper. 

1 W. F. Iibby, "Archaeology and radiocarbon dating", Radiocarbon 22 (1980), 1017-1020. 

2 J. R. Arnold & W. F. Libby, "Age determinations by radiocarbon content: checks 
with samples of known age", Science 110 (1949), 678-680. 

3 H. S. Smith, "Egypt and C14 dating", Antiquity 38 (1964), 32-37. 

4 T. Save-Soderbergh & I. U. Olsson, "C14 dating and Egyptian chronology", in: 
I. U. Olsson, ed., Radiocarbon variations and absolute chronology (Stockholm, 1970), 501-511. 


to achieve replication and inter-laboratory checks. But Save-Soderbergh 
and Olsson also noted the uncertainties attending the historical dates, 
especially those prior to the second millennium BC. 

Overall, Egyptian chronology contributed positively to the develop- 
ment of radiocarbon dating in the earlier decades: the apparent discrepan- 
cies observed between the radiocarbon age of some third millennium 
BC samples (mainly from Egypt) versus their 'known' age led to focussed 
interest in the investigation of such anomalies.'' Such work, using espe- 
cially known age tree-rings, led to the realisation that the relationship 
between radiocarbon and solar (calendar) years was neither equivalent 
nor fixed. 6 The development of increasingly accurate and precise records 
of such secular variation in natural radiocarbon levels became the dom- 
inant theme in radiocarbon dating for the next generation; already by 
the late 1960s to early 1970s calibration curves existed to convert radio- 
carbon years to calendar years back to beyond 5000 BC. 7 

The advent of calibrated radiocarbon dating, which had the effect 
of making many prehistoric contexts in Europe older than previously 
believed, had a radical impact in prehistoric archaeology — in particu- 
lar leading to the replacement of the previous 'diffusionist' models. 8 
Calibration also meant that the radiocarbon ages for Egyptian samples 
needed reconsideration, and a series of papers quickly addressed the 
radiocarbon dates from Egypt in the light of the initial proposals for 
an approximate calibration curve. 9 However, although the calibrated 
ages made general sense, the radiocarbon dates continued to be of 

5 W. F. Libby, "Accuracy of radiocarbon dates", Antiquity 37 (1963), 7-12. 

6 H. deVries, "Variation in the concentration of radiocarbon with time and location 
on Earth", Koninklyke Nederlandse Akademie van Wetenschappen (Series B) 61 (1958), 94-102. — 
H. E. Suess, "Secular variations of the cosmic-ray-produced carbon 14 in the atmos- 
phere and their interpretations", Journal of Geophysical Research 70 (1965), 5937-5952. 

7 E.g. H. E. Suess, "Bristiecone-pine calibration of radiocarbon time 5200 BC to 
present", in: Olsson (n. 4), 303-312. 

8 C. Renfrew, "The tree-ring calibration of radiocarbon: an archaeological evalua- 
tion", PPS 36 (1970), 280-311. — Idem, Before civilisation. The radiocarbon revolution and pre- 
historic Europe (London, 1973). 

9 I. E. S. Edwards, "Absolute dating from Egyptian records and comparison with 
carbon-14 dating", Philosophical Transactions of the Royal Society of London A 269 (1970), 
11-18. — R. M. Derricourt, "Radiocarbon chronology for Egypt and North Africa", 

JNES 30 (1971), 271-292.— R. M. Clark & C. Renfrew, "Tree-ring calibration of radio- 
carbon dates and the chronology of ancient Egypt", Nature 243 (1973), 265-270. — 
R. D. Long, "Ancient Egyptian chronology, radiocarbon dating and calibration", J^AS 
103 (1976), 30-48. — R. M. Clark, "Bristlecone pine and ancient Egypt: a re-appraisal", 
Archaeometry 20 (1978), 5-17. 


neither the accuracy nor precision to be of any real use to Egyptology; 
furthermore, the routine radiocarbon technology of the time required 
large sample sizes that were often problematic or impossible in terms 
of acquisition from archaeological excavations or from monuments or 

Over the subsequent quarter of a century radiocarbon dating has 
dramatically improved in terms of accuracy, precision, and sample size 
requirements. 10 In tandem, the natural and anthropogenic cycles and 
variations in atmospheric radiocarbon levels have become quantified in 
considerable detail. 11 The necessity of careful archaeological and other 
analysis to ascertain the security of association between the sample to 
be dated and the context for which a date is required is now fully 
appreciated (seminal paper by Waterbolk 197 1). 12 Programmes of inter- 
laboratory checking have greatly improved general standards in the 
field. 13 New technologies like accelerator mass spectrometry (AMS) per- 
mit dating of tiny samples, 14 and several routine radiocarbon labora- 
tories refined accuracy and precision to what is termed 'high-precision' 
level. Today the leading high-precision laboratories can demonstrate 
both good correspondence between measured ages and known real tree- 
ring ages, and good agreement between the laboratories, within the 
presently possible precision margins of c.2%o — that is within c. 10— 20 
radiocarbon years for the periods discussed in this paper. 15 The key 

10 R. E. Taylor, Radiocarbon dating, in: R. E. Taylor & M. J. Aitken, eds., Chronometric 
dating in archaeology (New York, 1997), 65-96. 

11 T. F. Braziunas, I. E. Fung & M. Stuiver, "The pre-industrial atmospheric 14 CC>2 
latitudinal gradient as related to exchanges among atmospheric, oceanic, and terres- 
trial reservoirs", Global Biochemical Cycles 9 (1995), 565-584. — M. Stuiver & T. F. Braziunas, 
"Anthropogenic and solar components of hemispheric 14 C", GRL 25 (1998), 329—332. — 
I. Levin & V. Hesshaimer, "Radiocarbon — a unique tracer of global carbon cycle 
dynamics", Radiocarbon 42 (2000), 69-80. 

12 H. T. Waterbolk, "Working with radiocarbon dates", PPS 37 (1971), 15-33. 

13 E.g. International Study Group, "An inter-laboratory comparison of radiocarbon 
measurements in tree-rings", Nature 298 (1982), 619-623. — E. M. Scott, A. Long & 
R. S. Kra, eds., "Proceedings of the international workshop on intercomparison of 
radiocarbon laboratories", Radiocarbon 32(3) (1990), 253-397. And further international 
inter-comparison exercises since. 

14 H. E. Gove, "The history of AMS, its advantages over decay counting: applica- 
tions and prospects", in: R. E. Taylor, A. Long & R. S. Kra, eds., Radiocarbon after 

four decades: an interdisciplinary perspective (New York, 1992), 214-229. 

15 E.g. M. Stuiver, P. J. Reimer, E. Bard, J. W. Beck, G. S. Burr, K. A. Hughen, 
B. Kromer, G. McCormac, J. van der Plicht & M. Spurk, "INTCAL98 radiocarbon 
age calibration, 24,000-0 cal BP", Radiocarbon 40 (1998), 1041-1083.— B. Kromer, 
S. W. Manning, P. I. Kuniholm, M. W. Newton, M. Spurk & I. Levin, "Regional 14 C0 2 
offsets in the troposphere: magnitude, mechanisms, and consequences", Science 294 


outcome of the latter development was the creation in the mid-1980s 
of a high-precision calibration of the radiocarbon timescale for the BC 
period. 16 

Shaw 17 quickly tried out high-precision calibration for existing Egyptian 
samples using the Irish Oak data of Pearson et a/. 18 He found the cal- 
ibrated ages to be in general agreement with the historical chronology, 
but did not see them as able to offer a useful alternative. Shaw was 
uncomfortable with the 'wiggles' in the calibration curve, and the sit- 
uation where a given radiocarbon age could yield two or more calen- 
dar age ranges. It was Hassan and Robinson 19 who finally brought 
methodological sophistication and chronometric hygiene to bear for 
Egyptian radiocarbon dates. They reanalysed the corpus of radiocar- 
bon data from Egypt against the 1986 high-precision calibration curve. 
They found that with suitable samples radiocarbon often could yield 
results compatible with the historical chronology; 20 and they highlighted 
the ability of radiocarbon to date directly a whole range of Egyptian 
contexts not closely tied into written records and the chronology of the 
pharaohs — a hint of the future real relevance of radiocarbon to (espe- 
cially non-elite) Egyptian archaeology and its chronology. But they also 
concluded that the existing corpus of radiocarbon data as of 1987 was 
not, with a few exceptions, fully satisfactory — they instead looked for- 
ward to better measurements in the future and then the fulfilment of 
the promise of radiocarbon dating. 21 

(2001), 2529-2532.— F. G. McCormac, A. G. Hogg, T. G. F. Higham, J. Lynch- 
Stieglitz, W. S. Broecker, M. G. L. Baillie, J. Palmer, L. Xiong, J. R. Pilcher, D. Brown 
& S. T. Hoper, "Temporal variation in the interhemispheric 14 C offset", GRL 25 (1998), 

16 G. W. Pearson, J. R. Pilcher & M. G. L. Baillie, "High precision 14 C measure- 
ments of Irish oaks to show the natural 14 C variations from 200BC to 4000BC", 
Radiocarbon 25 (1983), 179-186. — G. W. Pearson, & M. Stuiver, "High-precision cali- 
bration of the radiocarbon time scale", 500-2500 BC, Radiocarbon 28 (1986), 839-862. — 
M. Stuiver & G. W. Pearson, "High-precision calibration of the radiocarbon time scale, 
AD 1950-500 BC", Radiocarbon 28 (1986), 805-838. 

" I. M. Shaw, "Egyptian chronology and the Irish Oak calibration", JNES 44 (1985), 

18 Pearson et al. (n. 16). 

19 F. A. Hassan & S. W. Robinson, "High Precision radiocarbon chronometry of 
ancient Egypt, and comparisons with Nubia, Palestine and Mesopotamia", Antiquity 61 
(1987), 119-135. 

20 See also B. Weninger, "Theoretical radiocarbon discrepancies", in: D. A. Hardy 
& A. C. Renfrew, eds., Thera and the Aegean world III. 3: Chronology (London, 1990), 
216-231. — B. Weninger, Studien zur dendrochronologischen Kalibration von archdologischen 
"C-Daten (Bonn, 1997). 

21 Cf. (n. 19) at p. 129. 


But sadly there have been at best limited attempts to provide such 
better quality radiocarbon data for the periods after the Archaic — where 
increasingly good data exist. 22 Instead, publications by leading Egyptian 
chronological specialists concerned with the 3rd through earlier 1st mil- 
lennia BC in the late 1990s through 2003 largely dismissed or ignored 
radiocarbon evidence; 25 with Kitchen stating that '"science" cannot 
solve the intricate problems of detailed Egyptian successions, and the 
cross-links with the neighbouring Near East; texts alone can do that'. 24 
Such scholars cannot see any use for radiocarbon dating versus the 
believed-in dating accuracy and precision available from textual evidence. 
Ironically, the potential modern relevance of radiocarbon to Egyptology 
has been brought to the fore by a set of publications in the 1990s, 
which sought to question and reject the standard chronological syn- 
thesis and instead to propose a radically different (lower) Egyptian 
chronology for the second and earlier first millennia BC. 2) These writers 
appreciated that radiocarbon dating offered an independent check on 
their claims — they thus sought to dismiss or downplay radiocarbon dat- 
ing evidence. 26 In reality, however, radiocarbon evidence from the east 
Mediterranean indicated the reverse: that the range of the standard 
chronology was correct. 27 Radiocarbon perhaps had a use after all for 
Egyptian chronology. 

22 J. Gbrsdorf, G. Dreyer & U. Hartung, " 14 C dating results of the Archaic Royal 
Necropolis Umm el-Qaab at Abydos", MDAIK 54 (1998), 169-175.— S. H. Savage, 
"AMS 14 Carbon Dates from the Predynastic Egyptian Cemetery, N7000, at Naga-ed- 
Der", JAS 25 (1998), 235-249. Cf. A. R. Millard & T. A. H. Wilkinson, "Comment 
on 'AMS radiocarbon dates from the Predynastic Egyptian Cemetery, N7000, at Naga- 
ed-Der' by S. H. Savage", JAS 26 (1999), 339-341— S. H. Savage, "Towards an AMS 
radiocarbon chronology of Predynastic Egyptian ceramics", Radiocarbon 43 (2001), 

23 E.g. Beckerath, Chronologie, 55-56. 

24 K. A. Kitchen, "Ancient Egyptian chronology for Aegeanists", Mediterranean Archaeology 
and Archaeometry 2.2 (2002), 5-12 at p. 11. 

23 P. James, I. J. Thorpe, N. Kokkinos, R. Morkot & J. Frankish, Centuries of darkness 
(London, 1991). — D. M. Rohl, A test of time. 1: The Bible — -from myth to History (London, 

26 James et al. (n. 25), 321-325. — P.James, N. Kokkinos & I.J. Thorpe, "Mediterranean 
chronology in crisis", in: M. S. Balmuth & R. H. Tykot, Sardinian and Aegean chronol- 
ogy: towards the resolution of relative and absolute dating in the Mediterranean (Oxford: Studies 
in Sardinian Archaeology V, 1998), 29-43.— Rohl (n. 25), 384-387. 

27 E.g. S. W. Manning & B. Weninger, "A light in the dark: archaeological wiggle 
matching and the absolute chronology of the close of the Aegean Late Bronze Age", 
Antiquity 66 (1992), 636-663.— S. W. Manning, B. Weninger, A. K. South, B. Kling, 
P. I. Kuniholm, J. D. Muhly, S. Hadjisawas, D. A. Sewell & G. Cadogan, "Absolute 
age range of the Late Cypriot IIC period on Cyprus", Antiquity 75 (2001), 328-340. 


2. Radiocarbon Dating and the Historical Timescale 

The problem, historically, is that earlier radiocarbon dating offered at 
best large possible date ranges for any given measurement and these 
dates thus seemed an order of magnitude less accurate or precise than 
those available from the historical chronology of Egypt (the point made 
by von Beckerath). 28 And, as the review of Weinstein showed, 29 up until 
the early 1980s it is true that radiocarbon simply lacked the ability to 
supply the precision required in calendar years to be relevant to the 
existing, quite refined, historical chronology for the ancient Near East. 
But the advent of high -precision calibration curves from the mid-1980s, 
and increased accuracy and precision for standard radiocarbon dates, 
dramatically changed the situation. It was now possible to approach 
the precision of the historical chronology, and radiocarbon dating could 
thus offer an independent chronology free from the assumptions and 
step-wise logic transfers inherent in the existing chronological synthesis 
for Egypt and, there from, for the whole east Mediterranean. 30 

Recent developments emphasise this position. Following the first inter- 
nationally recommended high -precision calibration curves of 1986, a 
second internationally recommended extension, refinement, and revi- 
sion was made available in 1998 31 — and another (IntCal04) has been 
published while this text was in press. 32 Radiocarbon calibration datasets 
from tree-ring records from the east Mediterranean 33 have confirmed 
the local relevance of the standard northern hemisphere calibration for 
most periods (such work has inaddition identified some intervals of pos- 
sible regional/temporal variations for further study linked to key periods 
of short-term solar irradiance minima and climate change issues). 34 The 

28 Beckerath, Chronologie, 56. 

29 J. M. Weinstein, "Radiocarbon dating in the southern Levant", Radiocarbon 26 
(1984), 297-366. 

30 H. J. Bruins & W. G. Mook, "The need for a calibrated radiocarbon chronol- 
ogy of Near Eastern archaeology", Radiocarbon 31 (1989), 1019-1029. 

31 Stuiver el al. (n. 15). 

32 P. J. Reimer et al., "Radiocarbon calibration from 0-26 cal kyr BP", Radiocarbon 
46, 3 (2004), 1029-1058. 

33 Kromer et al. (n. 15). — S. W. Manning, B. Kromer, P. I. Kuniholm & M. W. 
Newton, "Anatolian tree-rings and a new chronology for the east Mediterranean Bronze- 
Iron Ages", Science 294 (2001), 2532-2535.— S. W. Manning, M. Barbetti, B. Kromer, 
P. I. Kuniholm, I. Levin, M. W. Newton & P. J. Reimer, "No systematic early bias 
to Mediterranean 14 C ages: radiocarbon measurements from tree-ring and air samples 
provide tight limits to age offsets", Radiocarbon 44(3) (2002), 739-754. 

34 See B. Kromer, M. Korfmann & P. Jablonka, "Heidelberg radiocarbon dates for 


development and application of stratified archaeological 'wiggle-match- 
ing' techniques have in turn allowed the exploitation of both (i) refined 
archaeological knowledge (stratigraphy) and (ii) the now refined and 
specific history of past natural atmospheric levels of radiocarbon entailed 
in these calibration curves in order to yield accurate and highly pre- 
cise calendar age ranges for sets of seriated samples. 3 ' Radiocarbon dat- 
ing has thus moved from a resolution of century-scale at best, to now 
being capable of decadal scale resolution. Radiocarbon can thus now 
have relevance at the 'historical' timescale. 

This does not mean that everything is now simple and clear; there 
remains plenty of scope for ambiguities and inconsistencies — as illus- 
trated in several of the papers in the recent Bruins et al. volume in 
Radiocarbon. ? ' b In particular, without selection of directly relevant sam- 
ples from primary contexts (e.g. short- to shorter-lived samples from 
secure and specific archaeological contexts relevant to the archaeological 
event/phase for which a date is sought), and then proper pre -treatment, 
and accurate and precise measurement in the laboratory, nothing use- 
ful will be gained. Old wood is clearly a major problem with some 
Egyptian samples 3 ' (see Section 4 below). The need for quality control 
at radiocarbon laboratories in terms of known-age blind checks is widely 
appreciated these days. Attention is increasingly moving now to the 
consideration of the integrity of the sample itself as offering only the 

Troia I to VIII and Kumtepe", in: G. A. Wagner, E. Pernicka & H.-P. Uerpmann, 
eds., Troia and the Troad: scientific approaches (Berlin, 2003), 52-53. 

35 E.g. Manning and Weninger (n. 27). — C. E. Buck, J. A. Christen & G. N.James, 
"BCal: an on-line Bayesian radiocarbon calibration tool", Internet Archaeology 7 (1999), 
[] — C. E. Buck, C. D. Litton & S.J. Shennan, 
"A case study in combining radiocarbon and archaeological information: the early 
Bronze-Age settlement of St. Veit-Klinglberg, Land Salzburg, Austria", Germania 72 
(1994), 427—447. — B. Weninger, "Stratified 14 C dates and ceramic chronologies: case 
studies for the Early Bronze Age at Troy (Turkey) and Ezero (Bulgaria)", Radiocarbon 
37 (1995), 443-456. — C. Bronk Ramsey, "Radiocarbon calibration and analysis of 
stratigraphy: the OxCal program", Radiocarbon 37 (1995), 425-430. — J. A. Christen & 
C. D. Litton, "A Bayesian approach to wiggle-matching", JAS 22 (1995), 719-725. — 
J. A. Zeidler, C. E. Buck & C. D. Litton, "The integration of archaeological phase 
information and radiocarbon results from the Jama River Valley, Ecuador: a Bayesian 
approach", Latin American Antiquity 9 (1998), 160-179. — C. Bronk Ramsey, J. van der 
Plicht & B. Weninger, "'Wiggle matching' radiocarbon dates", Radiocarbon 43 (2001), 

36 H. J. Bruins et al., "Near East Chronology: archaeology and environment", Radio- 
carbon 43 (2001), 1147-1390. 

37 M. Lehner, S. Nakhla, Z. Hawass, G. Bonani, W. Wolfli, H. Haas, R. Wenke, 
J. Nolan, W. Wetterstrom, "Dating the pyramids", Archaeology 52(5) (1999), 26-33. 


age of interest. Thus does the sample remain intact with only the radio- 
carbon age from the time the sample was exchanging with the atmos- 
phere, or have contaminating materials become included? And have 
there been processes of post-depositional diagenesis at work which are 
relevant? The need to investigate bone samples to ensure good colla- 
gen preservation is already appreciated and various strategies have been 
adopted. 3 " Although typically not likely to be a significant issue in 
general, the need to confirm removal of potentially contaminating humic 
material from archaeological wood/charcoal/seed samples should be a 
focus of further work. 39 

3. Radiocarbon and Egypt: An Example of Historically Relevant Data 

Integrated archaeological and radiocarbon analyses in other parts of 
the world carried out over the last decade have shown that, with high 
quality sampling and analysis, it is possible and practical to resolve 
chronology accurately and precisely down to the near-historical timescale. 
Although there has not yet been a significant body of work for Egypt 
after the Archaic period (e.g. refs. Section 1 above), it is important to 
appreciate that radiocarbon is now capable of offering relevant and 
independent dating for the OK through TIP. What is needed are mod- 
ern research programmes. To demonstrate that radiocarbon can poten- 
tially provide useful data which can either confirm and test historical 
chronology (where available), or can provide near-historical level dat- 
ing for those many other archaeological contexts in Egypt for which 
secure historical dates are not available, I review one example. The 
lack of my ability to note several good examples reflects the history of 
the field (cf. previous sections), and the failure so far to exploit radio- 
carbon where it could be most useful. 

38 R. E. M. Hedges, "Bone diagenesis: an overview of processes", Archaeometry 44 
(2002), 319-328. — G.J. Van Klinken, "Bone collagen quality indicators for palaeodi- 
etary and radiocarbon measurements", JAS 26 (1999), 687-695. 

39 See already D. Alon, G. Mintz, I. Cohen, S. Weiner & E. Boaretto, "The use 
of Raman spectroscopy to monitor the removal of humic substances from charcoal: 
quality control for 14 C dating of charcoal", Radiocarbon 44 (2002), 1—11. 


Tell el-Amarna 

Among existing radiocarbon dates from Egypt, one suite demands atten- 
tion. 40 These are five dates on a range of materials (bone, horn, skin, 
wood and charcoal) collected specifically and carefully for a high-quality 
programme of radiocarbon dating 41 from modern excavations at Tell 
el-Amarna, the short-lived capital of Egypt for most of Akhenaten's 
reign, founded in his 5th regnal year or ca. 1350/1346 BC. 42 The city's 
relative chronology is based on seventeen successive vintages documented 
in its epigraphical record, fourteen of Akhenaten himself (years 4 to 
17), and three belonging to his successors. 43 The city was deserted before 
the delivery of an eighteenth vintage. The specific context of the sam- 
ples taken for radiocarbon analysis was a midden probably deposited 
early within the site's history and thus it would date during the 13 
years of Akhenaten's reign at the site. 44 Hence the historical date range 
might be narrowed to between c. 1350/ 1346 BC to 1338/1334 BC. 

The Amarna radiocarbon ages on both known shorter-lived samples 
(skins, bone and horn) and on the potentially longer-lived wood and 
charcoal samples, offer a tight and coherent set of results entirely con- 
sistent with the historical dates and disprove any radically different 
chronology: Figure III. 1.1. We can see that the final interpretation of 
the radiocarbon data is very much determined by the shape of the 
radiocarbon calibration curve in the 14th— 13th centuries BC: see Figure 
III. 1.2. There is a sharp 'wiggle' upwards centred 1325 BC (confirmed 
for the east Mediterranean from Anatolian trees). 45 The Amarna data 
(bone sample Qj2505 perhaps apart) clearly do not match the peak of 
the wiggle, and thus could lie on either side. At 2o (95.4%) confidence, 
we see almost equal probability for either 1389-1329 BC or 1323-1260 
BC. The former range (and especially the most likely sub-range at lo 
confidence of 1373-1338 BC) matches the historical age estimate very 
closely. In support, we might note what seems to be an anomaly in 
the five-date set. The wood sample Q;2401 yields the second youngest 
(i.e. second most recent) radiocarbon age, and the animal bone sample 

40 Also Hassan & Robinson (n. 19), 123. 

41 V. R. Switsur in: Kemp, Amarna Reports I (1984), 178-188. 

42 Murnane & Van Siclen, Stelae, 73-86; Kitchen, "Chronology I", 39—52. — Kitchen 
(n. 24). 

43 Cf. above; Chapter II. 8. 

44 Cf. Switsur (n. 41), 181-182. 
43 Cf. n. 33. 




Boi indary _Bou id 

1 ) hase Tell El -Amarna 
(J-2401 3035±35BP 


Q-2404 302p±35BP 
O2405 308M?BE- 

Boi indary _Bou id 


C_Dale Amarna Cintext Earliest -1350 
C Dale Amarna CintextLatesl -1334 


2000BC 1800BC 1600BC 1400BC 1200BC 
Calibrated Date 

1000BC 800BC 

Sequence {A= 1 18.2%(A'<;= 60.0%)} 

Phase " 

j Tell E 
Q\2401 111. 


Q2402 110.9% 

Q-2403 111. 

Q-2404 108. 





2000BC 1800BC 1600BC 1400BC 1200BC 1000BC 800BC 
Calendar Date 

Figure III. 1.1. A. Calibrated calendar ages for the radiocarbon data reported from 
Tell el- Amarna, Egypt (Switsur, n. 41, 178-188) compared to the historical date for 
the context (see text). The upper and lower lines under each histogram indicate respec- 
tively the la (68.2%) and 2a (95.4%) calibrated age ranges. B. Sequence analysis (solid 
histograms) of the Amarna data (with the individual probabilities from A. indicated by 
the hollow histograms) as a phase within calculated boundaries. The Amarna data are 
entirely consistent with the historical age estimate for the context. Calibration and 
analysis employing OxCal 3.9 (Bronk Ramsey, n. 35 and later versions, with curve 
resolution set at 4) and INTCAL98 (n. 15). Q,-2401, wood; Q,-2402, charcoal; Q,-2403, 
skin; Q,-2404, horn; Q;2505, bone. Weighted average of all five data: 3050±16 BP (1), 
weighted average of just the three definitely shorter-lived samples 3054±20 BP (2). 2a 
(95.4%) confidence calibrated ranges respectively (1) 1388-1331 BC (46.6%), 1322-1260 
BC (48.8%), and (2) 1393-1260 BC (94%), 1228-1222 BC (1.4%). 



- 3200BP 

! 3150BP 

ST 3100BP 


I 3Q50BP 

§' 3000BP 

B 2950BP 


average : 305(M6BP 

68.2% probability 
1372BC(35.2%) 1338BC 
1318BC(22.6%) 1295BC 
1274BC(10.5%) 126fflC 

1388BC(46.9%) 1331BC 
1322BC(48.5%) 1260BC 

1500CalBC 1400CalBC 1300QilBC 1200Gj1BC 
Calibrated Date 


Figure III. 1 .2. Calibrated probability distribution for the weighted average radiocarbon 
age from the five measurements on samples at Tell el-Amarna reported by Switsur, 
n. 41). For discussion, see text. Calibration and analysis employing OxCal 3.9 (Bronk 
Ramsey, n. 35 and later versions, with curve resolution set at 4) and INTCAL98 

(n. 15). 

Q_-2405 yields the oldest age, and, in general, the average age of the 
likely longer-lived samples (wood and charcoal) at 3045±25BP, is (just) 
younger than the average age of the shorter-lived samples (animal skin, 
horn and bone) at 3054±20BP. Yet one would expect the wood sam- 
ple to be older in real calendar terms than the animal bone sample 
(by a few years or even a few decades or more). Out of the dating 
possibilities for each sample, the only way for this likely correct sam- 
ple relationship to occur is for the wood sample to date around the 
earliest of its three potential intercept ranges with the calibration curve 
at c. 1368-1 360 BC (and not c. 13 15-1 289 BC or c. 1280-1 262 BC), and 
for the animal bone sample to date around the later of its two possi- 
ble intercept ranges at c. 1336-1 320 BC (and not c. 1394-1 375 BC). 
And, plausibly, for the other three samples to date around or in between 
these preferred ranges. In turn, the mid to later 14th century BC date 
range is most likely for the Amarna samples. This is exactiy compatible 
with, and in support of, the standard Egyptian chronology, and, via 


the cuneiform text linkages attested at Amarna, 4(> this finding in turn 
supports and requires the standard Assyrian-Babylonian chronological 
range for this period. 47 Hence again radiocarbon provides useful inde- 
pendent support to Egyptian and ancient Near Eastern chronology, and 
disproves attempts to install radical chronological alternatives. 

4. Past Radiocarbon Fluctuations (the Shape of the Calibration Curve), the Old 
Wood Problem, and Egyptian OK Radiocarbon Dates 

A study by Haas et a/., 48 which indicated radiocarbon ages for vari- 
ous OK monuments several centuries earlier than expected, was widely 
seen as both a problem, 49 and by some as a good reason to avoid 
radiocarbon dating in Egyptology. The Haas et al. finding was largely 
repeated in the followup study by Bonani et al. 50 But it is not at all 
clear that there is any unknown 'problem'. A key issue is the history 
of past natural radiocarbon levels; there was in effect a plateau in radio- 
carbon levels in the period 2900—2500 BC. This means that radiocar- 
bon ages for the period 2900-2500 BC typically could intercept at 
several places with the radiocarbon calibration curve (i.e. several calendar 
periods have similar radiocarbon ages). For example, if we consider the 
OK monuments thought to be constructed c. 2600-2500 BC, then the 
wood employed will, at the latest, have its outermost ring dating then, 
and the rest of the relevant tree will be progressively older. Depending 
on species and source of the wood, one might expect an average offset 
of several decades to a century, give or take a range, for an average 
wood sample (e.g. compare the +50 ±50 old wood adjustment esti- 
mated by Vogel et al.')/' 1 Thus the 'average' wood used in a monument 

46 Summary in Beckerath, Chronologie NR, 23-24; cf. above, Chapter II. 13. 
'" J. A. Brinkman, Materials and studies for Kassite History (Chicago, 1976). 

48 H. Haas, J. Devine, R. Wenke, M. Lehner, W. Wolfli & G. Bonani, "Radiocarbon 
chronology and the historical calendar in Egypt", in: O. Aurenche, J. Evin & F. Hours, 
eds., Chronologies du Proche Orient/ Chronologies in the Near East. Relative chronologies and absolute 
chronology 16,000-4,000 B.P. (Oxford: BAR Int. Ser. 379, 1987), 585-606. 

49 Hassan & Robinson (n. 19), 129. 

50 G. Bonani, H. Haas, Z. Hawass, M. Lehner, S. Nakhla, J. Nolan, R. Wenke, 
W. Wolfli, "Radiocarbon dates of Old and Middle Kingdom monuments in Egypt", 
Radiocarbon 43 (2001), 1297-1320. 

51 J. S. Vogel, W. Cornell, D. E. Nelson &J. R. Southon, "Vesuvius/Avellino, one 
possible source of seventeenth century BC climatic disturbances", .Nature 344 (1990), 


built in the reign of Cheops in the mid or third quarter of the 26th 
century BC (conventional date ranges) likely dates during the first quar- 
ter of the 26th century BC give or take about 50 years — let us say 
2587+50 BC in broad terms. If we simulate the radiocarbon age, and 
its calibration, for 2587+50 BC, we get a result like that shown in 
Figure III. 1.3. And what we find is that the shape of the calibration 
curve (which represents the history of past natural variations in atmos- 
pheric radiocarbon levels) yields a calibrated age that seems 100—300 
years too old in the main and only just includes the real date at the 
very end of the calibrated range at 95.4% probability. But we cali- 
brated the 'correct' radiocarbon age! The point is that radiocarbon 
dating of single context events in this period is problematic because of 
the history of natural radiocarbon variations. Only use of another 
approach (like wiggle -matching) 52 can overcome this limitation. 

We can in fact generalise the potential and problems of OK radio- 
carbon dating by simulating radiocarbon ages for calendar years across 
this period. Figure III. 1.4 shows two runs of simulated dates at ±50 
years dating precision for the period 2750-2300 BC at 25 year inter- 
vals and including also the weighted average radiocarbon ages deter- 
mined and used for calibration by Bonani et al. 53 for the Pyramid of 
Snofru at Maidum and the Pyramids of Cheops, KJiephren and Mycerinus 
at Giza. Each run of a simulation produces different data from within 
the possible range. Thus note how the calibrated age for 2525 BC at 
+50 precision can vary quite a bit from a 'low' date range in Figure 
III. 1.4. A to a 'high' date range in Figure III. I.4.B. Samples near a 
slope in the calibration curve have more such potential for movement; 
other samples are much more stable. What we see is that the four sets 
of Dyn. 4 pyramid data lie entirely within the expected calibrated range 
for real dates from c.2750 BC to 2600 BC; they could be consistent 
with data from as late as c.2475 BC, but clearly prefer a date range 
starting around 2600 BC and older (compare also Figure III. 1.3 where 
the data want to lie on the plateau 2850—2600 BC and not so much 
on the slope following c.2600 BC). Such an outcome seems entirely 
plausible for the non-specific wood/charcoal samples (including 'flecks 

52 For examples at this time period, see e.g. B. Weninger, "Die Radiocarbondaten", 
in: M. Korfmann, ed., Demircihiiyuk: Die Ergebnisse der Ausgrabungen 1975—78. II Natur- 
wissenschqftliche Untersuchungen (Mainz, 1987), 4-13. — Weninger (n. 35). 

53 Cf. n. 50. 


of carbon in mortar' such as from the Cheops pyramid shown in Lehner 
et al.) 5 * from the Cheops, Khephren and Mycerinus monuments built 
in the 26th to early 25th century BC where average sample age is 
probably of the order of c. 50+50 years at the time of inclusion into 
the monument. (We therefore see that the radiocarbon 'dates' thus can 
be valid/correct — but they date the 'old' wood (etc), and not the cultural/ 
historical target date wanted: the building of the pyramid monument). 

The Pyramid of Snofru at Maidum 55 provided data where six of the 
seven dates are closely comparable — SMU-1412 on a 'log' is either 
aberrant or very old wood notwithstanding the stated dating of its 'outer 
rings' — and five of the determinations are stated to date outer rings 
from wood from the burial chamber (see Lehner et al.) 56 or shaft thereto. 
Thus these samples might be expected to derive from closer to the 
construction period of the monument (with this period usually assumed 
to start at year 2 of the reign, onwards). Following the 'historical' 
chronology, work on this monument began c.2600 BC (Stadelmann)" 
or 2638/2588 BC (Beckerath), 58 2616 BC (Kitchen) 59 or 2574 BC 
(Baines & Malek). 60 The calibrated age range of the average of these 
six similar 14 C ages given by Bonani et al. hi (2855—2583 BC at la, and 
2860-2579 BC at 2a) is entirely compatible at the end of its range (for 
why it will be just the end, see Figure III. 1.3 above) with the 'his- 
torical' age estimates (and especially not the lowest of these). The cal- 
ibrated probability distribution is entirely similar with a real age of 
c.2600 BC (see Figure III. 1.4). One may therefore conclude that the 
radiocarbon ages are approximately valid. 

The plateau in radiocarbon levels clearly creates difficulties for nar- 
row dating for OK samples. However, we may make some progress 
with current debates. For example, Spence proposed a rather lower 
OK chronology based on a hypothetical stellar alignment used by the 
pyramid builders. 62 She proposed dates of 2526±7 BC for the start of 

54 Lehner et al. (n. 37), 31 bottom left illustration 

55 Bonani (n. 50), 1304. 

56 Lehner et al. (n. 37), 31 top right illustration 

" R. Stadelmann, "Beitrage zur Geschichte des Alten Reiches. Die Lange der 
Regierung des Snofru", MDAIK 43 (1986), 229-240.— R. Stadelmann, Die Grqfien 
Pyramiden von Giza (Graz, 1990). 

58 Beckerath, Chronologie. 

59 Kitchen (n. 43). 

60 J. Baines & J. Malek, Atlas of ancient Egypt (Oxford, 1980). 

61 Bonani (n. 50). 

62 K. Spence, "Ancient Egyptian chronology and the astronomical orientation of 


work at the Snofru pyramid at Maidum, 2480+5 BC for the Cheops 
pyramid, 2448±5 BC for the Khephren pyramid and 2415±10 BC for 
the Mycerinus pyramid. However, if one examines Figure III. 1.4, it 
is evident that the radiocarbon data from these monuments are less 
consistent with such a very low chronology unless very, very old wood 
is always assumed. The range of simulated calibrated ages for samples 
dating 2450—2400 BC are not at all similar with the radiocarbon ages 
obtained from the Khephren and Mycerinus samples. The latter clearly 
date much older wood, wood preferably 100-150 years older. The 
more traditional range of 'historical' chronology estimates provides more 
suitable dates (allowing for a plausible average old wood factor where 
relevant; cf. also below, Chapter III. 4). 

Apart from the general calibration issue discussed above, some other 
issues may also be noted with regard to the Bonani et al. data sets. 63 
This team has published an enormous number of radiocarbon dates 
from OK and MK monuments. There are wide spreads of ages in sev- 
eral of the sets — suggested by the team involved themselves to be partly 
if not largely accounted for by an 'old wood' issue, as all available trees 
in the region, of widely varying ages, were consumed by the pyramid 
builders and as older settlement debris was recycled in fires. 64 While 
this is plausible in many cases, nonetheless, some samples are clearly 
aberrant for unspecified reasons. It is undoubtedly the case that the 
association of measured age for the sample (biological age unless other 
contaminating processes were involved) versus the date for monument 
construction is not demonstrated or clear in a number of instances (e.g. 
'charcoal' from mudbricks or from mortar — see Bonani et al. 65 — may 
easily represent 'old' or re -used tree-rings). The limestone and mortar 
associated with a number of samples may also provide a source of 
old carbon — for example with reference to samples from 'flecks of 
carbon in mortar' such as from the Cheops pyramid. 66 It is certainly 
interesting that the two secure datasets from early second millennium 
BC MK monuments (Pyramid of Senwosret II at Illahun and Pyramid 
of Amenemhet III at Dahshur), a new phase of pyramid building after 
a significant interval, yielded calibrated ages compatible with historical 

pyramids", Mature 408 (2000), 320-324. Cf. D. Rawlins, K. Pickering and K. Spence, 
"Astronomical orientation of the pyramids", Mature 412 (2001), 699-700. 

63 Cf. n. 50. 

64 Lehner et al. (n. 37), 33. 

65 Cf. n. 50, 1297-1298. 

6< ' Lehner et al. (n. 37), 31 bottom left illustration. 



32(X)QilBC 3000CalBC 2800CalBC 2600CalBC 2400QilBC 22(X)Ca]BC 
Calibrated Date 

Figure III. 1.3. Simulated radiocarbon age for a calendar date of 2587+50 BC using 
OxCal 3.9 (Bronk Ramsey, n. 35 and later versions with curve resolution set at 4) and 
INTCAL98 (n. 15). Note: every simulation run produces a slightly different outcome — 
this is a 'typical' output based on a number of runs. This simulated calendar age is an 
example of possible typical age of 'average' wood from a monument built in the reign 
of Cheops (reign starts 2604/2593/2554/2551 BC and ends 2581/2570/2531/2528 
BC: Beckerath, Chronologie; Kitchen n. 43; Baines & Malek, n. 60), given a typical +50 
±50 year 'old wood' adjustment. The calibrated radiocarbon age intersects with the 
c. 2900-2500 BC plateau in the radiocarbon calibration curve, and offers several lo 
ranges within a large 2a range c. 2880-2580 BC. We thus see as a function of the nat- 
ural history of radiocarbon fluctuations that real dates in the early 26th century BC 
yield calibrated radiocarbon ranges mainly apparently too early, with the real date just 
creeping into the last few years of the calibrated age range. It is noteworthy that the 
average radiocarbon age from the 45 samples used for the Pyramid of Cheops at Giza 
is in fact 4147±10BP (Bonani et al, n. 50, 1315) — almost exactly the 4154BP radiocarbon 
age derived by the simulation of a calendar date of 2587±50 BC, as shown above. Thus 
it appears that the Cheops data do, on average, yield a plausible age for wood used 
in his reign (with this average wood typically +50 or so years in age, give or take a 
range, versus the actual use date — and the large range within the Cheops data, see 
below, indicates such a range, or more, in the real wood ages, apart from any conta- 
minating factors from associated mortar/limestone). Because of the history of radiocar- 
bon variations (the 2900-2500 BC plateau), only the very last part of the calibrated 
range indicates the real age. These data, and the other similar OK data in Bonani et al. 
(n. 50) where the 'real' age at best creeps into the end of the calibrated age range, or 
lies shortly afterwards, therefore do not provide evidence of any additional offset beyond 
old wood and the calibration outcomes given the history of natural radiocarbon levels 
c. 2900-2500 BC. They provide no evidence at all for claims of hypothetical 100-300 
years too early offsets in Mediterranean radiocarbon ages based on claims of putative 
upwelling of old carbon (Keenan) 67 — something for which there is no positive evidence 
within an order of magnitude (Manning et al. 2002, n. 33). 

6/ D. J. Keenan, "Why early-historical radiocarbon dates downwind from the Medi- 
terranean are too early", Radiocarbon 44(1), (2002), 225-237. 


estimates. 68 The data from Archaic contexts also yielded radiocarbon 
ages largely in keeping with approximate 'historical' estimates 69 — as have 
other recent studies. 70 These periods both have helpful, non-plateau, 
radiocarbon calibration curve shapes, and may also plausibly have had 
less of an exhausted natural supply of wood — contrast the peak OK 
period of pyramid construction which probably forced much recycling 
of old material. 71 

For the third millennium BC, Bonani et al. report 17 date sets of 
the OK as older than their stated historical estimate, 6 as compatible, 
and 4 as more recent than the historical estimate. This clearly 'seems' 
to be a problem. But, apart from noting that the historical age esti- 
mate is commonly regarded as ±100 years for this period, the inter- 
pretation of Bonani et al. is based on two inappropriate starting points. 
First, there is no allowance for likely average sample age at time of 
use (i.e. 'old-wood' age for random wood/charcoal samples not known 
to be outer tree-rings), and second Bonani et al. use average values for 
the radiocarbon age of sample sets which contain significant internal 
variation, and this is thus potentially misleading (probably less so as set 
size increases). To illustrate: examination of Bonani et al. (n. 50, Fig. 1) 
shows the Cheops Pyramid (object number 13) to yield one of the 
apparently tighter calibrated age ranges and to be some two centuries 
older than the estimated historical age.' 2 And this despite 46 radio- 
carbon dates being reported for the monument. 73 But examination of 
the 46 radiocarbon data reveals ages varying by 1210 radiocarbon 
years! — and even excluding the two gross outliers in the set, 74 the age 
range left in the set is still 513 14 C years! As shown in Figure III. 1.5, 
just over one-third of the individual samples — the younger ages — do in 
fact offer calibrated ages more or less compatible with the estimated 
historical age of 2589—2566 BC, 75 and most of the remainder offer 

68 Bonani (n. 50), Fig. 1 and p. 1320. 

1,9 Bonani (n. 50), Fig. 1 object numbers 1—5. 

70 E.g. Gorsdorf et al. (n. 22). 

" Lehner et al. (n. 37). 

12 Bonani et al. (n. 50), 1315 use the historical range of 2589-2566 BC — other stan- 
dard sources suggest c.2593-2570 BC: Kitchen (n. 43), 48; around 2604/2554 BC to 
2581/2531 BC: Beckerath, Chronologie, 188, or 2551-2528 BC: M. Lehner, The com- 
plete pyramids (London, 1997), 8. 

73 Bonani (n. 50), 1305. 

'* Marked by the * and + signs by Bonani (n. 50), 1305. 

75 Bonani (n. 50), 1316. 




3000CalBC 2500CalBC 

Calibrated Date 


Figure III. 1.4 (A and B). Two outputs of simulations of calibrated radiocarbon ages 
for calendar years 2750-2300 BC at +50 dating precision and at 25 calendar year 
intervals (data from OxCal 3.9 with curve resolution set at 4, Bronk Ramsey, n. 35 
and INTCAL98, n. 15). Included also are the weighted average radiocarbon ages used 
for calibration by Bonani et ai, n. 50, 1314-1316) for the Pyramid of Snofru at 
Maidum, the Pyramid of Cheops at Giza, the Pyramid of Khephren at Giza and the 
Pyramid of Mycerinus at Giza. For discussion, see text. 


R Simulate 

2675BC -2675±50 
2650BC -2650±50- 
2625BC -2625±SQ- 


Snefru at Msydum 41 1023BP 

R Simulate 

2600BC -2600±5Q- 

R Simulate 


2575BC -2575±50 

2550BC -2550±50 

R Simulate 


2500BC -2500±50 
Giza 4127±25BP 
2475BC -2475i50 

R Simulate 

2450BC -2450±50 

R Simulate 

2425BC -2425±50 

R Simulate 

2400BC -2400±50 
2375BC -2375i50 
2350BC -235O±50 
2325BC -2325i50 

R Simulate 

2300BC -2300±50 

3500CalBC 3000CalBC 2500CalBC 

Calibrated Date 



variously a little to quite a bit older ages — 'old wood" 6 — would appear 
the obvious first hypothesis. 77 Such a pattern: younger ages corresponding 
to, or close to, context date and older ages reflecting old wood is quite 
common and expected when dealing with wood/charcoal samples. 78 
Similar observations may be made about the data sets for: Step Pyramid 
of Djoser at Saqqara, Temple Complex associated with the Step Pyramid, 
Pyramid of Sekhem-khet at Saqqara, 79 Pyramid of Khephren at Giza, 
Pyramid of Ra c djedef at Abu Rawash, Sphinx Temple of Khephren 
at Giza (n. 50, 1306), Pyramid of Mycerinus at Giza, Mortuary Temple 
of Shepseskaf at South Saqqara (n. 50, 1307), Mortuary Temple and 
Pyramid of Sahure at Abusir (n. 50, 1309) and Pyramid of Teti at 
Saqqara (n. 50, 1310). In contrast, it is notable that the radiocarbon 
ages from a modern excavation at the Royal Production Centre at 
Giza offer both a reasonably consistent set, and calibrated ages more 
recent than the surrounding OK datasets from the monuments. K() 

We have already noted the case of the Pyramid of Snofru at Maidum, 
where six of the determinations date outer rings from wood from the 
burial chamber or shaft thereto. And thus the usual old-wood effect is 
likely minimised. The calibrated age range of the average of these six 
similar 14 C ages given by Bonani et al. (2855—2583 BC at la, and 
2860-2579 BC at 2a) is entirely compatible at the end of its range (for 
why it will be just the end, see Figure III. 1.3 above) with the 'his- 
torical' age estimate employed by Bonani et al. (n. 50, 1314) or those 
estimated by Kitchen (n. 43) or Beckerath, Chronologie, (higher range), 
and also overlaps with the date for the accession of Snofru c.2600 BC 
and his earlier reign, and thus the construction of this Snofru's first (of 
three) pyramids, given by Stadelmann (n. 57). One may observe that 
the stated calibrated range ends +8/+4 years from the start of the 
lower 'historical' age estimate for Snofru by Lehner from Baines & 
Malek (n. 60); K1 this is hardly a significant difference, and the wood in 

76 M. B. Schiffer, "Radiocarbon dating and the 'old wood' problem: the case of the 
Hohokam chronology", JAS 13 (1986), 13-30. 

77 See Lehner (n. 37), esp. 31-33. 

78 For an example from Troy II, see Kromer et al. (n. 34), 48 and Fig. 4. 

79 Bonani (n. 50), 1303. 

80 Bonani (n. 50), Fig. 1 object 12, contrasted with other objects 10-19. One might 
speculate that the samples from this context, which are not from major architecture/ 
monuments and their creation industries, do not therefore suffer so much from an 
average old wood problem. 

81 Lehner (n. 72). 


question could easily have been cut a few years earlier than the start 
of Snofru's reign if the lower dates are to be preferred, just as it could 
have been cut during his reign if the slightly higher dates are preferred. 
We have already seen how even the correct radiocarbon age for a date 
around the early 26th century BC only includes the real calendar age 
within the very end of the calibrated range, as much of the dating 
probability ends up on the plateau in the radiocarbon curve over the 
preceding couple of centuries (compare Figure III. 1.3 above). Thus 
these dates for the Pyramid of Snofru at Maidum are entirely consis- 
tent with the estimated 'historical' age. 

In sum, these OK radiocarbon dates do not in fact indicate any 
problem with radiocarbon dating and Egyptian chronology; instead they 
nicely illustrate the importance and impact of the shape of the cali- 
bration curve in dating, and they highlight the need to obtain organic 
samples directly associated with, and relevant to, the human context 
for which a date is sought. Wood and charcoal samples especially can 
easily be older, or even much older, than their final deposition con- 
text depending on tree species and the uses and perhaps re -uses of the 
wood. Add in calibration taphonomy and correct radiocarbon ages for 
organic materials can appear to yield dates that are centuries too old 
for their historical/archaeological context (see Figure III. 1.3). Aquatic 
samples, which may include a water/marine reservoir radiocarbon age 
(versus solely the normal atmospheric reservoir radiocarbon age repre- 
sented in normal terrestrial plants, and animals eating these), must also 
be treated with care and caution — again this may explain some of the 
apparently aberrant radiocarbon ages obtained on 'reed' samples in 

As evident from the Amarna example in Section 3, and other stu- 
dies cited in Sections 1 and 2, or other studies in the literature," 2 in 
appropriate circumstances high-quality radiocarbon data from Egypt 
and the east Mediterranean region can provide accurate and precise 
dates). 83 When issues occur, such as the completeness of removal of age 
contamination by humic material, 84 or the old wood offsets evident in 
the extensive OK radiocarbon measurements published by Bonani et al. 

82 M. Lange, "Wadi Shaw 82/52: 14 C dates from a peridynastic site in northwest 
Sudan, supporting the Egyptian historical chronology", Radiocarbon 40 (1998), 687-692. 

83 See also Manning et al. (2002, n. 33). 

84 Alon et al. (n. 39). 




Boundary _Bound 

Phase Khufu 













ETH-13762 3927±58BP 

ETH-13754 3980±57BP 

ETH-13756 4143±61BP 

ETH-13757 4225±79BP 

ETH-1376' 3928±54BP 
ETH-13763 3937±61BP 
ETH-0307 4440±90BP 
ETH4227 4215±105BP 
ETH-13770 4087±53BP 
ETH-1377' 4187±60BP 
ETH-1377Ji 4190±52BP 

ETH-13777 4313±57BP 

6000BC 5000BC 4000BC 3000BC 

Calibrated Date 



Figure III. 1.5 (A and B). Calibrated calendar ages for the 46 radiocarbon data reported 
from the Pyramid of Cheops at Giza (data from Bonani et al. n. 50, 1305). The esti- 
mated historical age employed by Bonani et al. (n. 50, 1315) is 2589-2566 BC. The 
upper and lower lines under each histogram indicate respectively the lo (68.2%) and 
2o (95.4%) calibrated age ranges. Calibration employing OxCal 3.9 (Bronk Ramsey, 
n. 35 and later versions, with curve resolution set at 4) and INTCAL98 (n. 15). For 
discussion, see text. 



3 hase Khufu 

Phase ETH-1 3778 4156±58B 

ETH-1 3779 4062±61BP 

ETH-0308 4300±85BP 

ETh+4228 4390±110BP 

ETH-1 3783 4237±62BP 

ETH-1 3784 4068±54BP 


ETH-1 3782^ 

ETH-1 37£lf 

ETH-1 3791 

ETH-0309 4420±10()BP 


ETH-1 3800 4195±55BP 

ETH-1 3799 4128±58BP 

ETH-1 3801 4189±60BP 

ETH-1 3802 4174±61BP 

ETH-1 3803 
ETH-1 3805 


— I 1 1 1 \— 



ETH-0334 4440± 320BP 

ETH-1 3900 4068±60BP 


^ETH-0313 4330±125BP 
Boundary _ESound 

7000BC 6000BC 5000BC 4000BC 3000BC 2000BC 1000BC 

Calibrated Date 


(n. 50), they in fact lead us to consider important topics concerning 
taphonomy, sample diagenesis, social history, economic processes, and 
the environment. As Lehner et al. conclude with regard to the OK 

If the fair agreement of our 1995 results with historical dates and previ- 
ous radiocarbon dates for the Archaic period and with the historical dates 
for the MK hold, the problematic OK dates are boxed in. And therein 
may lie a hint of multifaceted old wood effects for a period, especially 
from Djoser to Mycerinus, when any and all wood resources may have 
been consumed at a whole other order of magnitude than before or after 
the giant pyramid-building projects . . . our project . . . now has us think- 
ing about forest ecologies, site formation processes, and ancient industry 
and its environmental impact . . . 

5. Caution: The Need to Make Only Secure Historical Associations is 

Radiocarbon dating determines the age of an organic sample. The asso- 
ciation of such a sample and its radiocarbon age with history/ archaeology 
is the task of the archaeologist. And one has to be careful and rigor- 
ous. Associations must be demonstrated, not casually assumed. A recent 
example illustrates the potential problems and the need to be even 
more careful as better precision becomes available in modern radio- 
carbon dating. 

Shoshenq I and Radiocarbon? 

A key synchronism for the standard chronology of Egypt (and wider 
Near Eastern history) concerns the identification of the important Egyp- 
tian pharaoh Shoshenq I with the Shishak attested in the Bible (I Kings 
14:25—26; II Chronicles 12:3—4) as invading Judah and Israel in the 
5th year of Rehabeam. B(> Rehabeam year 5 is in turn dated c. 926/925 
BC by linking the attested names and reign lengths of the 10th— 9th 
century BC kings of Israel and Judah with recorded synchronisms in 
the 9th century BC between the Israelite kings Ahab and Jehu and the 

85 Lehner (n. 37), 33. 

86 Kitchen, TIP, esp. xliv, 72-76, 287-302; Kitchen (n. 24), 7-8; Beckerath, Chronologie 
MR, 30-34; idem, Chronologie, 68-70. 


Assyrian king Shalmaneser III. Since the chronology of the Assyrian 
kings is effectively absolute back to the 10th century BC, this enables 
precise calendar dates to be applied (of respectively c.853 BC for the 
last year of Ahab's reign and c.841 BC for the first year of Jehu's 

In an important and controversial paper, Bruins et al. (2003) recently 
reported sets of high-precision radiocarbon dates, allied with an inter- 
pretative stratified archaeological wiggle-matching analysis, from the site 
of Tel Rehov in Israel. 87 These samples, on high-quality short-lived 
samples, provide the basis for a high-resolution chronology for the site 
in the 12th through 9th centuries BC. But Bruins et al. also suggested 
that the date for the destruction of Stratum V at Tel Rehov could be 
associated with the campaign of Shoshenq I and thus their date for 
this stratum — c. 940-900 BC — was argued to support this proposed 
identification, and in turn the standard Egyptian chronology or one 
very close to it (with the Shoshenq I invasion dated c. 926/925 BC — 
see above, or various slight alternative calculations, such as the 918 BC 
of Miller and Hayes," 8 cited by Bruins et al. (n. 35) 317, or 927 BC in 
Barnes 89 or 922-921 BC in Hayes and Hooker. 90 

However, the critical logical step was missing. There is no evidence 
at all that the Stratum V destruction links with Shoshenq I — this is 
merely an unproven and (unnecessary) assumption incorporated into 
the Bruins et al. paper and its dating model (and so leads to a circu- 
lar argument). The dating of the site and the dating of Shoshenq I 
are separate until and unless clear evidence can be produced to show 
that Shoshenq I caused the specific Stratum V destruction horizon 
dated by the radiocarbon measurements. Archaeologists must always 
be aware that non-rigorous and specific (i.e. documented) assumptions 
that try to bring archaeological and historical evidence together (the 
event-historical model) are often inherently problematic because the 
respective evidence types represent fundamentally different facets of 

87 Bruins et al. n. 35. 

88 M. Miller & Hayes, A history of ancient Israel and Judah (Philadelphia, 1986). 

89 W. H. Barnes, Studies in the chronology of the divided monarchy of Israel (Cambridge, 
Mass., 1991). 

90 J. H. Hayes & P. K. Hooker, A new chronology for the kings of Israel and Judah and 
its implications for Biblical history and literature (Atlanta, 1988). 


historical reality. 91 In the Shoshenq I case, it is fair to note that much 
is less than certain and very different narratives are possible based on 
the same limited and likely non-contemporary 'historical' evidence/ 
tradition. 92 

This example highlights the need to delineate clearly what is the tar- 
get date and how and why the relevant organic samples do (or do not) 
provide associated dating evidence when radiocarbon dated. Without 
such chronometric care results are not credible, and conclusions may 
turn out to rest on foundations of sand. 

6. Conclusions 

High quality radiocarbon dating offers an important but as yet not fully 
exploited resource for Egyptology. It provides an increasingly accurate 
and precise test for the historical chronology and can actively inform 
and resolve disputes in less certain or ambiguous periods. Available dat- 
ing accuracy and precision from radiocarbon should in principle — i.e. 
on suitable short-lived samples from primary contexts dated at good 
precision — offer a chronological precision for the third millennium BC 
of the order of, or better than, the historical age estimates — which, for 
this period, are often regarded as having a significant error margin of 
up to a century. It could test and resolve claims for significantly differing 

91 A. M. Snodgrass, "Archaeology", in: M. Crawford, ed., Sources for ancient history 
(Cambridge, 1983), 137-184. — Idem, An archaeology of Greece: the present state and future 
scope of a discipline (Berkeley, 1987), 37-66. 

92 I. Finkelstein, "The campaign of Shoshenq I to Palestine: a guide to the 10th 
century BCE polity", Z Dp V 118 (2002), 109-135. 

Postscript. Since the present text was submitted in August 2003, there have been sev- 
eral further publications (and much discussion) taking this topic now well beyond the 
initial publication of Bruins et al. (2003) cited above. However, the logic/methodology 
point noted in the text remains relevant as outlined (and has since been accepted by 
the Bruins et al. authorship — I wish to thank Hendrik Bruins and Amihai Mazar for 
friendly, constructive, and productive discussion). For the latest (AD 2005) situation on 
the analysis of the important Tel Rehov datasets, see now (i) Mazar, A. et al. "Ladder 
of Time at Tel Rehov: Stratigraphy, Archaeological Context, Pottery and Radiocarbon 
Dates", and (ii) Bruins, H. J. et al. "The Groningen Radiocarbon Series from Tel 
Rehov: OxCal Bayesian Computations for the Iron IB-IIA Boundary and Iron IIA 
Destruction Events", both papers in Radiocarbon Dating and the Iron Age of the Southern 
Levant — the Bible and Archaeology Today, edited by Thomas Levy and Thomas Higham, 
Equinox Publishing, Ltd., London (2005), 193-255, and 271-293 respectively. 


dates, such as those suggested from speculative astronomical conjecture 
by Spence (n. 62). 

For the second millennium BC radiocarbon may also be able to 
assist. For many non-elite contexts it may offer the best means of dat- 
ing. For the chronology of the pharaohs it will be of less need, as the 
dates in the second millennium BC are relatively accurately and pre- 
cisely determined from a combination of so-called 'dead-reckoning' (the 
compilation of documented names of Egyptian kings and various other 
persons and attested years of reign/office backwards from an agreed 
starting point fixed against the Greco-Roman timescale) 93 and analysis 
of some records of astronomical observations (e.g. Krauss; 94 Beckerath). 9:> 
Recent scepticism, and claims to reject for example all lunar data 
(Wells 96 — approvingly cited by e.g. Kitchen), 97 have been shown to 
be based on incorrect or partial understanding of the data and their 
analysis. 98 

The leading scholars immersed in the details argue that this combi- 
nation of historical data and astronomical evidence forms a closely 
dated chronological system for the second millennium BC, with only 
at most a few years to a decade or so error range, and with several 
likely absolute placements therein, such as the accession of Tuthmosis 
III in 1479 BC. 99 Nonetheless, one role for radiocarbon will be to offer 
an independent check and verification of these chronologies. Without 
this, complete certainty will never be possible given that there are gaps 
and uncertainties/ambiguities in the evidence (textual or astronomical), 
and key assumptions/interpretations have been made by modern schol- 
ars. Radiocarbon dating is direct and independent, and can cut through 
circular debates and assumptions. 

93 Kitchen, TIP, idem, "Chronology I"; idem n. 24. 

94 Krauss, Sothis. 

95 Beckerath, Chronologic, 41—51. 

96 R. A. Wells, "The role of astronomical techniques in ancient Egyptian chronology: 
the use of lunar month lengths in absolute dating", in: J. M. Steele & A. Imhausen, 
eds., Under one sky: astronomy and mathematics in the ancient Near East (Miinster, 2002), 

97 Kitchen, n. 24 at p. 11. 

98 R. Krauss, "Ronald A. Wells on astronomical techniques in Ancient Egyptian 
Chronology", DE 57 (2003), 51-56. 

99 Kitchen, TIP, n. 43; n. 24; Beckerath, Chronologic NR; 1997; cf. below, Chapter 
III. 8. 


High-quality radiocarbon dating also offers the independent means 
to test and reject the several publications of the last decade which have 
argued that conventional Egyptian (and wider ancient Near Eastern) 
historical chronology is incorrect. 100 

Radiocarbon dating should become the friend of Egyptologists. Whereas 
in its origins Egyptology helped to test radiocarbon dating and to expose 
the need for calibration, modern radiocarbon dating now offers the 
means to test, support, extend, and even to refine Egyptian chronol- 
ogy. Certain periods like the OK will be problematic if samples or con- 
texts are dated in isolation thanks to the unhelpful plateau in the 
calibration curve (see Figure III. 1.3 above); but by exploiting tech- 
niques like seriated archaeological wiggle-matching (ideally of short-lived 
samples tied securely to the context for which a date is sought), even 
this time period can be made to yield an accurate and precise cali- 
brated radiocarbon chronology by taking advantage of the shape of the 
calibration curve. 101 Radiocarbon dating also offers the route to engage 
with all those many Egyptian archaeological contexts not specifically 
linked with the (largely elite centred) textual record (compare the sim- 
ilar but still largely unfulfilled hope expressed twenty years ago by 
O'Connor. 102 The entirety of Egyptian archaeology can then be inte- 
grated into an accurate and precise near-historical level timeframe. 

100 E.g. J. Goldberg, "Centuries of darkness and Egyptian chronology: another look", 
DE 33 (1995), 11-32. — G. Hagens, "A critical review of dead-reckoning from the 21st 
Dynasty", JARCE 33 (1996), 153-163.— James et al. (n. 25), (n. 26).— Rohl n. 25.— 
P. Van der Veen & W. Zerbst, Biblische Archaologie am Scheideweg? (Holzgerlingen, 2002). 

101 See e.g. Weninger (n. 52); Weninger (n. 35). 

102 D. O'Connor, "New Kingdom and Third Intermediate Period", in: B. G. Trigger, 
B. J. Kemp, D. O'Connor & A. B. Lloyd, Ancient Egypt: a social history (Cambridge, 
1983), 183-278, esp. 185. 














V lntCal04+/-1o 


V\ lntCal98+/-1o 





— i — i — i — i — i — i — i — 1 — i — i — i — i — 1 — i — i — i — i — 1 — i — i — i — i — 1 — i — i — i — i — 1 — i — i — i — i — 1 








Calendar Date BC 

Figure III. 1.6. The new (AD 2005) IntCal04 radiocarbon calibration curve (black) at 
la for period 500 BC to 3500 BC, compared to the previous IntCal98 curve (grey) 
as used in this paper. There is little significant difference — the main change is that the 
IntCal04 curve is a little more smoothed. Data from Reimer et al. (n. 32) and Stuiver 

el al. (n. 15). 


Christian Goedicke 

The phenomenon of luminescence in solids has been used as a dating 
method in archaeology since the late 1960s. Since then the method 
developed as an important and universal tool alongside radiocarbon 
dating and dendrochronology. It is based on the property of a num- 
ber of minerals to store and release radioactive decay energy. Two 
minerals frequently occurring in archaeological contexts, quartz and 
feldspar, show this effect very distinctiy which makes the application of 
luminescence analysis in archaeology particularly useful. In the field of 
the geological sciences luminescence dating has caused a decisive increase 
of knowledge and is regarded meanwhile as indispensable. In the fol- 
lowing it will be discussed in which parts of archaeology the use of the 
method may be inadequate. 

Luminescence Dating: Basic Principles 

Radioactive traces (U-238, U-235, Th-232, K-40, Rb-87) are constituent 
in clay, in soil, and in rocks. During spontaneous decay, these elements 
release energy into the environment. When quartz or feldspar occurs 
in the environment, the emitted energy may be stored in the crystal 
lattice of these minerals. After an archaeological storage time the stored 
energy can be released in form of light. The longer the archaeological 
storage time, the larger the accumulated energy (phys. the dose) and 
the stronger the light signal. Luminescence dating is a dosimetric dating 
method requiring no external calibration. The necessary zero-setting 
event is achieved by heating or exposure to light when the energy 
stored in the lattice over geological times is zeroed and the accumulation 
can begin again. 

Due to different ranges and to different effectiveness in producing 
luminescence, the three types of radioactive decay radiation (alpha-, 
beta- and gamma radiation) need to be taken into account separately. 
Consequently, more than one physical quantity must be measured in 


the laboratory and, hence, the number of measurements results in a 
complex error of the age which is one of the major differences from 
C-14 dating. The total error of a thermoluminescence single date of 
an object amounts to between 7 to 12% of the total age expressed in 
years. In absolute numbers this amounts to roughly ±450 years for an 
OK date. The range of uncertainly is much smaller, if several samples 
from the same archaeological context are analyzed; the so-called con- 
text error can be reduced to approx. ±5% which still corresponds to 
±230 years for the OK. 

Thermoluminescence ages are calculated according to the following 

archaeological dose 

age = ; (Eq. 1) 

a-dose-rate„ + dose-rate« + environmental dose-rate 

internal dose-rate external dose-rate 

A feature peculiar to Egyptian artefacts is the low radioactive trace- 
element content which explains the low internal dose-rates frequendy 
encountered in Egyptian artefacts. Consequently, thermoluminescence 
ages become dependent on the ratio of the internal to external dose- 
rate (see Eq. 1). Should the two dose-rates come close to or equal each 
other, the external dose-rate may determine the thermoluminescence 
age. Fig. III. 2.1 illustrates the effect of the dose-rate ratio on the ther- 
moluminescence age. 

Thermoluminescence produces the best results when applied to arte- 
facts as they are excavated, i.e. when the measurement of the envi- 
ronmental dose-rate can be made on-site. No access to the site limits 
the usefulness of the method, e.g. analysis of museum objects of unknown 
provenance cannot produce definitive results. The humidity of a sam- 
ple also requires access to the site, as humidity is a correcting factor 
for the age of a sample. Variation of the humidity during storage times 
has to be taken into account. 



Fig. III. 2.1. Effect of the environmental dose-rate on thermoluminescence age; the 
example shows a typical ceramic sample from the OK. The fraction of the environmental 
dose-rate most probably prevailing in Egypt lies between ca. 15 and 30. 

Luminescence techniques can be used to date: 

A. fired materials: ceramics (minimum firing temperature 500°C), stones 
(firesides), casting cores of bronzes 

B. unfired materials: sediments (aeolian, fluvial, colluvial), mortar 

How Useful is Luminescence Dating for Dynastic 
Based on Published Dates 

ian Objects? Answers 

Considering the usual error margins of 7—12%, thermoluminescence 
dating of Egyptian artefacts cannot contribute much to chronological 
evidence. The method may prove useful for a piece which cannot be 
attributed to any period on stylistic criteria. Even the range of context 
errors exceeds the possible imprecision of the Egyptian chronology 
which in its present state is better than 5% in all periods. Hence it is 
not surprising that published examples of luminescence dating for 
Egyptian artefacts are few. Recent culling of various bibliographical 
sources including CAS (chemical abstracts services) turned up the fol- 
lowing papers. 


Thermoluminescence Dates Quoted in the Literature 

Two nearly identical papers deal with pottery of the fourth dynasty. 1 
The instrumentation used and the reported error of 2.3% cast some 
doubt on the seriousness of both; the standard calculation will not result 
in an error that small. 

A French team notes that the pigment Egyptian Blue results from 
a thermal reaction thus making dating of this material possible. 2 
The attempt to date a large chunk of Egyptian Blue from Karnak is 
described resulting in a reasonable date. However, dating pigments 
from paint-layers is not feasible due to dosimetry problems arising from 
paint-layer thickness which usually is smaller than the range of beta- 

Some unpublished results of the author (Rathgen-Forschungslabor in 
Berlin) on potsherds from excavations at Abu Minshat Omar can be 
mentioned here. A comparison of the results with radiocarbon data 
from the same site revealed a systematic deviation towards younger 

Samples of vitrified silt from the Early Dynastic tombs at Abydos 
turned out to be undatable. Secondary firing resulted in a high degree 
of vitrification making luminescence dating impossible. 

For earlier periods with a comparatively fluid chronological frame- 
work luminescence dates, even including the above quoted errors, may 
furnish a valuable chronological contribution. However, even for this 
period luminescence has rarely been used. Whittle's early study dealt 
with potsherds from Hemamieh and Qurna-Tarif. 3 Environmental 
dose-rate data were obtained in the laboratory by alpha-counting of 
soil samples. An archaeological evaluation of the data obtained was 
not attempted. Two additional studies considered Middle Palaeolithic 

1 S. A. Elfiki, M. S. Abdelwahab, N. Elfarawamy & M. A. Elfiki, "Dating of ancient 
Egyptian pottery using thermoluminescence techniques", Nuclear Instruments and Methods 
94 (1994), 91-94.— M. S. Abdel-Wahab, S. A. Elfiki, M. A. Elfiki, M. Gomaa, S. Abdel- 
Kariem & N. El-Faramawy, "Annual dose measurements and TL-dating of ancient 
egyptian pottery", Radiation Physics and Chemistry 47 (1996), 697-700. 

2 M. Schvoerer, M. C. Delavergne & R. Chapoulie, "The thermoluminescence of 
Egyptian Blue", .Nuclear Tracks and Radiation Measurements 14 (1988), 321-327. 

3 E. H. Whittle, "Thermoluminescent dating of Egyptian pre-dynastic pottery from 
Hemamieh and Qurna-Tarif", Archaeometry 17 (1975), 119-122. 


settlements. 4 Dating Neolithic cultures by thermoluminescence and radio- 
carbon in the Sudan Nile valley was the aim of a French working 


Luminescence techniques other than thermoluminescence, e.g. optically 
stimulated luminescence (OSL), are not very likely to improve the 
situation; conversely, the sensitive dependence on the dose-rate ratio 
will become even more important. Corresponding to the zero-setting 
event (bleaching by sun-/daylight) these techniques are more appro- 
priate for use in geological sciences. About half a dozen studies devoted 
to sediment dating and shoreline development have been done. 6 

4 N. Mercier, H. Valladas, L. Froget, J. L. Joron, P. M. Vermeersch, P. Van Peer 
& J. Moeyersons, "Thermoluminescence dating of a Middle Palaeolithic occupation at 
Sodmein Cave, Red Sea Mountains (Egypt)", Journal of Archaeological Science 26 (1999), 
1339-1345.— P. M. Vermeersch, E. Paulissen, S. Stokes, C. Charlier, P. Van Peer, 
C. Stringer & W. Lindsay, "A Middle Palaeolithic burial of a modern human at 
Taramsa HiU, Egypt", Antiquity 72 (1998), 475-484. 

5 P. Guibert, C. Ney & M. Schvoerer, "Datation croisee TL/C14 de cultures neo- 
lithiques de la Vallee du Nil, Soudan, sites d'El Kadada et d'El Ghaba", Archeologie du 
MIMoyen 5 (1991), 129-141.— P. Guibert, C. Ney, F. Bechtel, M. Schvoerer & F. Geus, 
"TL and radiocarbon dating of neolithic sepultures from Sudan: intercomparison of 
results", Radiation Measurements 23 (1994), 393-398. 

6 O. Bubenzer & A. Hilgers, "Luminescence dating of Playa Sediments of the 
Egyptian Plateau, Western Desert, Egypt", Quartemary Science Reviews 22 (2003), 1077- 
1084. — H. M. El-Asmar & P. Wood, "Quaternary shoreline development: the north- 
western coast of Egypt", Quaternary Science Reviews 19 (2000), 1137-1149. 


Otto Cichocki 


Dendrochronology is the science of dating tree rings. It includes inves- 
tigations of the information provided by/contained in the structures of 
dated ring sequences and its applications to environmental and histor- 
ical problems/questions. 1 


The astronomer Andrew S. Douglass expected to find growth reactions 
to the sun-spot cycle, when he measured the width of tree rings of old 
pine trees. What he found was a similar pattern of ring width com- 
mon to trees growing in the same area at the same time. He applied 
this discovery to develop a method to date the remains of pueblo settle- 
ments in the American Southwest. 2 

Subsequently the method was continually improved, especially by 
computer based statistical analysis in laboratories worldwide. 

Biological Background 

In regions with seasonal variation the growth of trees and other woody 
plants is not an invariable process but influenced by climatic factors. 
Particularly outside the tropical and subtropical zones, growth stops 

1 M. Kaennel & F. Schweingruber, Multilingual Glossary of Dendrochronology. Terms and 
Definitions in English, German, French, Spanish, Italian, Portugese and Russian (Bern, 1995), 91 

2 A. E. Douglass, "The secret of the southwest solved by talkative tree rings", National 
Geographic Magazine 54 (1929), 737-770. Idem, "Dating Pueblo Bonito and other ruins 
of the southwest", National Geographic Society. Contributed technical papers. Pueblo Bonito series 
(Washington, 1935), 1-74. 


almost completely in winter, causing a distinct border between each 
annual increment. Depending on genus and species, these "tree-rings" 
are composed of different cells in different arrangements. Early wood 
is springwood formed at the beginning of the growing season for con- 
ducting liquids, while late wood grows in late summer and autumn and 
is composed mainly of cells with thicker walls. Both together form a 
tree-ring. The thickness of the ring is more or less correlated with 
precipitation (if this is the determining factor), with density corresponding 
to the average temperature of the growth season. Of course, other cli- 
matic factors, but also insect damage, wounding or the climatic situa- 
tion of previous years, affect growth as well. Therefore each tree ring 
is the result of a very complex accumulation of influences. 

Trees growing in tropical and subtropical zones also form layers of 
different cell arrangement, but since they lack distinct borders, these 
layers cannot be measured for dating purposes. As annual thickness 
growth is the result of the reaction of trees to ecological influences, in 
an area with similar influences the growth reactions of different trees 
on comparable stands in the same year will be similar. 

Sampling, Data Acquisition, and Synchronisation 

Old wood may be preserved in different conditions: 

Dry preservation: wood is almost unaltered, but sometimes fragile. 
It can be found in buildings (used to reinforce walls, construct ceilings, 
roofs, lintels, or doors), in burials (coffins, other wooden objects, burial 
chambers), as the ground of a painting (icons, Fayum portraits), sculp- 
tures and other objects. 

Wet-wood preservation: wood is almost unaltered, but very sensitive 
to drying (necessitating storage in water, in a cool and dark environ- 
ment). It can be found in rivers, lakes, caves, at the bottom of wells 
(often together with other organic remains of importance). 

Charcoal preservation: burning changes wood — it is chemically resis- 
tant, but very fragile. If thicker beams are visible, they should be exca- 
vated separately, bandaged and parts plastered together and packed 
in plastic but not sealed to allow a very slow drying. Cool and dark 
storage is best. 

To measure the thickness of rings in a particular piece of wood, a 
cross-section is required. A disc can be cut from wooden architectural 
elements or a core removed with a drill. After smoothing the surface, 


the ring borders become visible. In most cases mixing samples of different 
wood species is not recommended when constructing a standard, so 
wood species analysis (identification of anatomical structures of a very 
small sample with the help of a light microscope) of all objects avail- 
able for investigation should be the initial step. 

If the surface (square or longitudinal) of a wooden object is not cov- 
ered with paint and has been well smoothed when made in antiquity, 
the rings can often be measured directly after preliminary cleaning. 
Measuring devices have been developed for this purpose since most 
items in museum collections are only available for non-invasive "on the 
object" measurements. 3 VideoTimeTable equipment uses a digital video 
camera with macro lens, which is moved along the surface by a step 
motor. A live video image, displayed on a laptop, can be measured 
immediately. Plain surfaces can be investigated with the help of a 
modified high-resolution flatbed scanner. A device capable of drilling 
a 5mm diameter hole for endoscopic measurement of ring width was 
designed to measure objects with paint or other surface treatment. 

At least two radii of a sample must be measured. The arithmetical 
mean value of the ring widths is calculated to compensate for biolog- 
ical diversity (e.g. elliptic ring shape, single growth deviations). The 
result is a list of mean growth-ring thickness for each year contained 
in the sample. A sample must contain at least 70 rings; otherwise in 
many cases the statistical methods used for dating cannot work. Exceptions 
to this rule are separate pieces of a single stem (e.g. charcoal), collected 
as a unit. In this case much shorter overlaps allow reconstructing the 
whole ring sequence of the stem because of almost identical ring pat- 
terns. It is necessary to collect and measure as many promising sam- 
ples as possible from one complex (a "time-unit") in order to calculate 
a reliable mean value list for further analysis. 

Constructing Standards 

Two samples grown at the same time in the same climatic conditions 
will show a statistically significant correlation of their distribution of 
wide and thin rings (respectively their mean values). Such tree-ring 

B. Knibbe & O. Cichocki, "Developing new tools for the SCIEM 2000 project" 
in press. 


patterns are significant for a certain time period, and a certain climatic 
region, but in many cases only for a single genus (or species) of trees. 
Cross dating (the statistical comparison of two mean value lists) aims 
at the synchronization of different samples and determining relative 
time span between them. 

If two samples do not completely overlap in time, their mean pat- 
tern will be longer than one of the individual samples. Many success- 
fully cross-dated overlapping samples from different periods — the youngest 
one being of known age (e.g. a recently cut tree) — allow constructing 
a tree ring standard (a dated chronology) valid for one species and a 
limited growth area. If a gap is open (a sample is missing), the isolated 
part is called a floating chronology. 

Dating of Samples 

Dating a sample means comparing a sample with a standard on a year- 
to-year basis to find the best matching position. In many cases statis- 
tical tests result in more than one possible date. To find the correct 
position, "pointer years" and additional statistical tests are used. When 
setting up a standard by cross-dating all contained samples, certain 
years form pronounced peaks in the graph. If more than 75 percent 
of the rings from one year have the same trend (increasing or decreas- 
ing growth compared with the preceding ring), this year with especially 
strong influence on growth is called a "pointer year" and is specially 
marked. In dubious cases, in a second match, only these pointer years 
are compared with the corresponding rings of the sample. The correct 
position will then show a significantly higher correlation of these spe- 
cial years than the other theoretically positions. 

The skeleton plot is a special method for dry areas using minimum 
pointer years. Here, only very narrow rings are observed, as they are 
formed by drought, that makes moisture the factor limiting growth. In 
years with poor environmental conditions some species cannot create 
distinct rings in the entire circumference of a stem or branch. If these 
missing rings go undetected, the sample will not match well with the 
standard. If this occurs when the standard is being set up, results will 
be negatively effected. Detection is possible by comparing many sam- 
ples, because the same ring is not usually missing in all of them. The 
best way to detect incomplete rings is by comparing different radii on 
cut discs. 


Another misleading growth pattern may occur in years with a tem- 
porary period of low temperature or drought during the growing sea- 
son which produces a "false ring". This "late wood formation" within 
the ring has a much smoother outer border than the real late wood 
formed at the end of the growth season. Hence, it can be detected 
under the microscope. As climatic factors are the reason certain rings 
are formed, it is possible to extract a basic part of this information 
contained in the ring. Dendroclimatology is concerned with recon- 
structing precipitation and temperature as major climatic factors. 4 

Wood grows annual layer upon annual layer. To avoid confusion, 
the age of a sample is usually defined as the absolute age of the 
outermost ring preserved in the sample. If this ring is the last grown 
before the death of the tree (the so-called waney edge, in some cases 
preserved with bark), it is possible to date the cutting year and to iden- 
tify the cutting season. Problems can arise from samples of wood in 
secondary use or from undetected repairs, for they may give a false 
higher or lower age, respectively, for their context. 

Dendrochronology in Egypt and the Near East 

The wood species to be investigated depend on their occurrence in 
the objects available for investigation and their suitability for this 
special method. Wood species analyses on Egyptian objects have been 
carried out on coffins and other objects in the British Museum, 5 on 
statues, wooden toilette objects, musical instruments and objects from 
the Coptic period in the Louvre, 6 and on various objects in the Munich 

4 F. H. Schweingruber, Tree Rings. Basics and Application of Dendrochronology (Dordrecht, 

' W. V. Davies, "Ancient Egyptian Timber Imports. An Analysis of Wooden Coffins 
in the British Museum", in: W. V. Davies & L. Schofield, eds., Egypt, the Aegean and 
the Levant. Interconnections in the Second Millennium BC (London, 1995). 

6 A. Nibbi, Ancient Egypt and Some Eastern Neighbours (New Jersey, 1981). 

7 D. Grosser et al., "Holz — ein wichtiger Werkstoff im Alten Agypten", in: S. Schoske 
et al., eds., Anch — Blumen fur das Leben. Schriften au.s der dgyptischen Sammlung, Heft 6, 
Staadiche Sammlung agyptischer Kunst (Miinchen, 1992). 


Table III. 3.1 

Type of Wood Nibbi 1981 Grosser 1992 Davies 1995 

Louvre Munich and others British Museum 

Ficus sycomorus 32 

Tamarix sp. > 100 

Acacia sp. 35 

Cedrus sp. 5 (7) 

Juniperus sp. 2 

Pinus sp. 9 
Cupressus sp. 















Table III. 3.1 provides a rough idea of the range of wood species used 
in ancient Egypt for various purposes. Unfortunately the majority of 
these species grow in Egypt (Ficus sycomorus, Tamarix sp., Acacia sp.) 
and since they do not form distinct ring borders, they are unsuitable 
for dendrochronological analysis. Only the gymnosperms can be used: 
Cedrus libani grows in Mediterranean mountain climate and has distinct 
ring borders. Its growth age is said to be as much as 500 years. Due 
to over-deforestation there are only a few small areas where it grows 
today in Lebanon, larger ones in Turkey (Taurus, Antitaurus), and in 
Syria. The sub-species Cedrus libani atlantka grows in the Adas Mountains 
of Algeria and Morocco, Cedrus libani brevifolia in a very limited area in 
the mountains of the island of Cyprus. There are no anatomical features 
that distinguish these three sub-species of cedrus. 

Cedar was imported to Egypt from the time of the Old Kingdom. 
Different interpretations of Egyptian texts mentioning imports of wood 
from certain countries, continue to fuel a heated debate about the origin 
of cedar found in Egypt. 8 Another contentious issue is the correct trans- 
lation of different hieroglyphic terms obviously characterizing different 
species of wood. 

Nili Lipschitz carried out wood species analysis and dendrochrono- 
logical investigations on several historical and archaeological sites in 
Israel. 9 She worked with wood of Cedrus libani, Cupressus semper- 

8 Davies (n. 5); R. Meiggs, Trees and Timber in the Ancient Mediterranean World (Oxford 
1982); Grosser et al. (n. 7); Nibbi (n. 6); idem, "Some Remarks on the Lexicon Entry: 
Zeder, Cedar", DE 7 (1987), 13-27. 

9 N. Lipschitz, "Overview of the Dendrochronological and Dendroarchaeological 
Research in Israel", Dendrochronologia 4 (1986), 37-58. 


virens, Pinus nigra, Quercus cerris and Pistacia khinjuk. Dating the 
roof beams of the El-Aqsa mosque in Jerusalem produced a 231 year 
floating chronology with 14-C dates of the Byzantine period. 10 

The Aegean Dendrochronology Project (directed by Peter I. Kuni- 
holm) is working on many archaeological sites in Turkey, Greece, and 
Italy. In the International Treering Data Base, standards for the 2nd 
millennium AD from Turkish forests are published for Cedrus libani, 
Juniperus sp., Quercus conferta, Pinus nigra, Pinus sylvestris and Abies 
nordmanniana. ' ' A floating chronology for Bronze Age and Iron Age 
by 14-C wiggle matching is based on data for different tree species. 1213 

Fritz Schweingruber published several modern standards for Cedrus 
libani brevifolia from Cyprus in the International Treering Data Base. 
A dendrochronological attempt at dating for Egypt continues/is car- 
ried out within the long-term scientific project "The Synchronization 
of Civilizations in the Eastern Mediterranean in the Second Millennium 
BC" at the dendrolab of VIAS, University of Vienna. The dendro- 
chronological investigations try to develop a dating approach indepen- 
dent of other archaeological and scientific methods. 14 First results are 
a 507 year floating Cedrus libani chronology for the 2nd millennium 
BC, and several shorter floating chronologies. To date this chronology 
includes/subsumes the coffin of Sebekhetepi, the inner coffin of Gua, 
and a canopic box (British Museum, London); the garden model of 
Meketre (Metropolitan Museum of Art, New York); the coffins of Ashait, 
Djehutinakht, Khnumhotep, Neferi, the outer coffins of Amenemhat 
and Mesehti, and the shrine of Auibre Hor (Egyptian Museum, Cairo). 

Wood samples from modern trees (Besharre, Barouk, Horsh Ehden) 
and from several buildings from Deir el Kamar and Qadisha valley 
(Lebanon) cover the centuries back to 1369 AD for further construc- 
tion of a standard for absolute dating. Analysis of Bronze age charcoal 

10 S. Lev-Yadun, "The Origin of the Cedar Beams from Al-Aqsa Mosque: Botanical, 
Historical and Archaeological Evidence", Levant 24 (1992), 201-208. 


12 P. I. Kuniholm, "A Date-List for Bronze Age and Iron Age Monuments based 
on combined Dendrochronological and Radiocarbon Evidence", in: M. J. Mellink, ed., 
Aspects of Art and Iconography: Anatolia and Its Neighbors. Studies in Honor of Nimet Ozgii( 
(Ankara, 1993), 371-373. Idem et al., "Anatolian tree rings and the absolute chronol- 
ogy of the eastern Mediterranean, 2220-718 BC", Nature 381 (1996), 780-783. 

13 S. Manning et al., "Anatolian Tree Rings and a New Chronology for the East 
Mediterranean Bronze-Iron Ages", Science 294 (2001), 2532-2535. 

11 O. Cichocki, "Cedres libanais comme instrument de datation en Egypte" 
of the 31CAANE Conference 2002; in press. 


samples from Arqa (Lebanon), Qatna, and Ebla (Syria) will enable test- 
ing of the standard for the 2nd millenium BC with 14C and should 
aid in determining whether Cedar wood was imported from Mount 
Lebanon or from other forests. 

Dendroclimatology studies the impact of climate on trees through 
patterns of growth. As different factors (moisture, temperature, length 
of vegetation period) interact in a very complex system, the investi- 
gated parameters (ring width, early wood/late wood ratio, density) con- 
tain a varying mean of signals. Investigations in the Near East carried 
out by a team lead by Ramzi Touchan reconstructed precipitation for 
southern Jordan. Currently this project has expanded to take in Turkey, 
Lebanon, and Syria. 1 ' Knowledge about climate impact on ancient 
economies may aid analyses of political crises or warfare, which in turn 
influenced wealth and long-distance trade. 

15 R. Touchan, D. Meko & M. K. Hughes, "A 396- Year Reconstruction of Precipi- 
tation in Southern Jordan", Journal of the American Water Resources Association 35(1) (1998), 


Rolf Krauss 

The rising and falling of the Nile and the harvesting of grain and fruit 
occur regularly at certain times within the solar year. When their dates 
were recorded in terms of the Egyptian calendar, they can be con- 
verted into absolute dates. Occasions of quarrying expeditions have also 
been traditionally considered examples of seasonal dates, since Egyptologists 
conjectured that such work was not undertaken during the summer. 1 
Even if so, the conversions yield intervals that are too broad to be of 
much use for chronology. For if cool weather lasts for 120 days, from 
November to February, then a "cool season date" spans an interval of 
4 x 120 years = 480 years. Although there might have been a ten- 
dency to send expeditions to quarries during the cooler months, there 
are nevertheless attestations for expeditions at the hottest time of the 
year. 2 Thus conversion of expedition dates can result in chronological 

Dates of the Mile Flood 

The Nile flood results principally from monsoon rains that fall over the 
Ethiopian plateau between mid-May and September. 3 In modern times 
the dams constructed at Aswan beginning around 1900 have prevented 
the annual flooding of the Nile Valley. Data recorded in the 19th cen- 
tury and the Middle Ages provide information about the onset and 
duration of the flood which are crucial for evaluating pharaonic dates. 
The Nile sunk to its lowest level in April/May; towards the end of 

1 Meyer, Chronologic, 177-178; idem, Nachtrage, 20; C.J. Eyre, "Work and organiza- 
tion of work in the Old Kingdom", in: M. A. Powell, ed., Labor in the Ancient Near East 
(New Haven, Connecticut, 1987), 16. 

2 Beckerath, Chronologic, 53. 

:i W. Willcocks, Egyptian Irrigation (London, 1889), 10. 


May and the beginning of June it began to rise. The biography of 
Weni, a Dyn. 6 official, seems to mention a low water date. 4 Weni 
reports that he brought an altar from Hatnub to the pyramid of Merenre c 
within 17 days in III Shemu, "although there was no water on the 
tzw". Lieblein, who dated Merenre c far too early at ca. 2525 BC, 5 cal- 
culated that III Shemu 17 corresponded to March 5 Greg., 6 a time 
when the river falls fast. Gardiner paraphrased Weni's description with 
"when the river was at its lowest".' If so, Merenre c would have reigned 
between ca. 2817 and 2694 BC when III Shemu 30 coincided with 
May 3 1 (Greg.) and III Shemu 1 with April 1 (Greg.). Following 
Gardiner and using the standard chronology for the OK, Eyre cor- 
rectly calculated the Gregorian months December and January as cor- 
responding to III Shemu, 8 but he wrongly designated these months as 
"a time of low, if not the lowest, water". 9 There are various possibili- 
ties to resolve the contradictions in Weni's report: the translation of 
tzw as sandbanks may be wrong; 10 the flood might have been low and 
run off very early that year; Weni may even have exaggerated. 

Another low water date refers to a difficult passage through the chan- 
nels of the Semna rapids at Uronarti in 19 Senwosret III. 11 Correlated 
with the Dal inscription of year 10 of Senwosret III, the Uronarti 
inscription "provides evidence for an extraordinary variability in the 
Nile levels of late winter during the reign of Senwosret III". 12 Seasonal 
dates that diverge from the statistical mean cannot be used to estab- 
lish absolute chronology. In the following paragraphs the basic data 
are presumed to be samples of the statistical mean. 

64 Gregorian maximum flood dates on record from the Middle Ages 
and the 19th and early 20th centuries refer to the Nile gauge at Roda 
(Old Cairo), whereas the pharaonic high flood dates refer to Karnak 
temple. At Roda the maximum height was reached between September 

4 Urk. I, 108.— Meyer, Chronologie, 178; Nachtrage, 20 n. 1. 

5 J. Lieblein, "Eine chronologische Bestimmung", £45 44 (1907), 101-102.— Cf. 
Borchardt, Mittel, 89 n. 3, for a similar approach. 

6 The correct correspondance of III Shemu 1 7 in 2525 BC is March 7 Greg. 
' Gardiner, Egypt, 97. 

8 If Merenre' reigned in ca. 2222 + 6 BC, then III Shemu corresponded in his 
reign to ca. December 8 to January 7 (Greg.). 

9 Eyre (n. 1), 16. 

10 Cf. J. Vandier, La Famine dans I'Egypte ancienne (Cairo: RAPH 7, 1936), 74-77. 

11 Borchardt, Mittel, 91. 

12 B. Bell, 47^1 79 (1975), 238. 


25 and October 5. 13 The earliest maximum date is August 14 and the 
latest October 27. Using Borchardt's incomplete Nile flood data, Beckerath 
presumed that the maximum occurred at Roda as early as August 25 
and as late as October 27, but 4 days earlier at Luxor. 14 For the lat- 
ter, Borchardt relied on the fact that in 1925 the maximum travelled 
from Luxor to Roda within 4 days. 15 However, the speed differed con- 
siderably in other years: "The Nile at Assuan reaches its maximum 
about the 5th September, and would under ordinary circumstances be 
at its highest in Lower Egypt about the 11th September, but as the 
basins of Upper Egypt are being filled in August and September, and 
emptied in October, the maximum in Lower Egypt is ordinarily about 
the 10th October." lb The delay apparently resulted from interaction 
between the filling of the irrigation basins and the size of the flood. 
For example, in 1887 the Nile reached its maximum at Aswan on 
September 1, at Armant on September 6, and at Roda on September 
25. I7 In that case 19 days elapsed between maxima at Luxor and Roda, 
instead of 4 days as Borchardt had presumed. 

Application of the 1 9th century flood data for Roda to the pharaonic 
period presupposes a comparable irrigation system. Irrigation works of 
some kind are first attested in the FIP, 1H but very little information sur- 
vives for later pharaonic history. There is no indication that a fully 
developed system of basin irrigation existed in the NK or even earlier. 
Under these circumstances it is preferable to use the maximum flood 
dates of Aswan, instead of Roda, when calculating dates for the Luxor 
region. The maximum required was at most 4 days for the distance 
Aswan: Luxor. 19 There are 34 maximum dates for Aswan on record; 20 
the earliest is August 18, the latest October 1, yielding maximum dates 
for Luxor between August 21/22 and October 4/5. Based on a com- 
parison of the dates at Aswan and Roda, it follows that the maximum 

13 W. Popper, The Cairo Milometer (Berkeley, 1951), 87-88. 

Borchardt, Annalen, 7; Borchardt, Mittel, 90; Beckerath, Chronologie, 52. 

Borchardt, Mittel, 90 n. 5. 

Willcocks (n. 3), 10. 

Willcocks (n. 3), 31, 184-185. 

E. Endesfelder, "Zur Frage der Bewasserung im pharaonischen Agypten", ZAS 
106 (1979), 37-51; W. Schenkel, Die Bewdsserungsrevolution im Alten Agypten (Mainz, 1978), 

19 For varying velocities of the Nile, see W. Willcocks, Egyptian Irrigation' 2 (London, 
1913), 143-144. 

20 H. G. Lyons, The Physiography of the River Mile and its Basin (Cairo, 1906), 289-290. 


gauge occurred between 4 days (1882) and 63 days (1894) at Roda 
later than at Aswan. 21 Borchardt's Roda-based maximum dates for 
Luxor tend to be much too late, yielding an interval of ca. 4 x 64 
years = 256 years for a pharaonic maximum date. By contrast, the 
Aswan-based interval for maximum dates at Luxor results in a period 
of ca. 4 x 45 years =180 years. Records for only three events at high 
flood are preserved: 

(1) Sebekhotep VIII, Epagomenae of year 4: flooding of Karnak temple. 22 

(2) Osorkon III, year 3, III Peret 22: 23 flooding of Karnak temple. 

(3) Shebitku, year 3, I Shemu 5: royal visit to Karnak temple after a 
high flood. 24 

(1) Provided Sebekhotep VIII ruled ca. 1600±150 years, then August 
22 (Greg.) corresponds to September 5 (Jul.) and October 5 (Greg.) to 
October 19 (Jul.). These Julian dates correspond to the E(pagomenae) 
as follows: 

E 1 = Oct 19 Jul in E 1 = Sept 5 Jul. in 
1709/06 BC 1532/29 BC 

E 5 = Oct 19 Jul in E 5 = Sept 5 Jul. in 
1693/90 BC 1516/13 BC 

The flood in 4 Sobekhotep VIII occurred, then, between 1709 and 
1513 BC; the mean year is 1611 BC. By contrast, Beckerath obtained 
the limits 1869 and 1534 BC and the mean year 1701 BC. 

(2) Provided Osorkon III ruled around 700 BC, then August 22 
(Greg.) corresponds to August 29 (Jul.) and October 5 (Greg.) to October 
13 (Jul.). The Julian dates for III peret 22 are: 

III peret 22 = October 13 (Jul.) in 861/858 BC 
III Peret 22 = Aug 29 in 681/678 BC 

21 For the maxmimum dates at Roda see Lyons (n. 20) 321 and M. Clerget, Le 
Carre I (Cairo, 1934), 44 (correct "1880 aout" to "1880 septembre"). 

22 L. Habachi, SAK 1 (1974), 207-214; J. Baines, Acta Orientalia 36 (1974), 39-54; 
idem, Acta Orientalia "il (1976), 11-20. 

23 Reading after Schott; see Borchardt, Mittel, 91. 

24 J.v. Beckerath, JARCE 5 (1966), 53; idem, GM 136 (1993), 7; idem, 
92, 52. 


Therefore the flood in 3 Osorkon III occurred between 861 and 678 
BC; the mean year being 770 BC. 2> By contrast, Beckerath obtained 
the limits 932 and 677 BC and 805 BC as the mean year. 

(3) The text that relates to Shebitku's appearance in Karnak men- 
tions the Nile level "of his time (m ffw.ff' '. Borchardt interpreted the 
visit as coincidental with a maximum flood whose occurrence was slightly 
delayed. 26 He conceded that the maximum flood could have occurred 
before the king's visit. I Shemu 5 corresponded to October 9 (Greg.) 
in 705 BC = 3 Shebitku. 2 ' It cannot be excluded that the visit coin- 
cided with a much delayed maximum flood, although more probably 
the maximum occurred earlier, in September. 

Opening of the Basin Canals in the Theban Area? 

Ramesside graffiti record dates when "this day, the water returned/came 
to inundation" 28 which were thought to refer to the falling of the flood 
or the feast "night of the drop" or the onset of the flood. 29 More 
recently Janssen believes the dates to refer to the opening of the basin 
canals that took place in the 19th century around August 12. He sup- 
poses that "the situation in Pharaonic times was not essentially different 
than in the nineteenth century AD," and that the Ramesside flood 
dates also fell around August 12 (Greg.). 30 Beckerath presumes that the 
dams were opened depending on the height of the flood, with the mean 
August 12 (Greg.). 31 Both authors overlook the recorded indiviual dates 
for the opening of basin canals in the 19th century. In Kena province, 
and thus at Luxor, the dates deviated from August 12 by +3 days and 
-6 days at most. 32 August 12 was a deadline set by the Khedive that 

25 Borchardt, Mittel 91 n. 6, noted that the day coincided with a procession of Amun 
(line 5 of the text); he expected a full moon. He may have been correct coinciden- 
talfy, if the Tepi Shemu date I Shemu 6, year 18 (Fitzwilliam 68d) is correctly ascribed 
to Osorkon III as Kruchten, Annates, 144, 240 suggests. 

26 Borchardt, Mittel, 91. 

27 Cf. above Jansen-Winkeln, Chapter II. 10, and Zibelius-Chen, Chapter II. 12. 

28 J. J. Janssen, JNES 46 (1987), 125-136. 

29 Meyer, Nachtrage, 39-42; with additional literature. 

30 Janssen (n. 28). 

31 Beckerath, Chronologie, 52. 

32 Willcocks (n. 19), 335. 


could be missed if a harvest of sorghum was imminent in a basin. 33 
The interpretations of Janssen and Beckerath presuppose that the sys- 
tem of basin irrigation existed already in Ramesside times and was 
managed as in the 19th century. No NK sources support this idea. 
The Ramesside flood dates may refer instead to the flooding of the 
valley after the Nile had breached the levees. 

Grain Harvesting Dates 

Ploughing and sowing were done as soon as the flood had receded. 
According to one MK source, the fields were measured for tax assess- 
ment when the seed had sprouted, 34 whereas NK sources seem to indi- 
cate the measuring of ripening fields. 33 Thus the measuring of fields 
cannot be used to establish absolute chronology. 

Traditionally, Egyptologists assumed that barley was harvested between 
the end of February and the beginning of March (Greg.), and wheat 
in April (Greg.). 36 By contrast, in pharaonic times the harvest could 
have began as early as February and ended as late as in May (Greg.). 3/ 
The delivery of grain occurred later, after time-consuming threshing, 38 
winnowing, and cleaning. 39 Most references to collection and delivery 
of grain are Ramesside and date to summer and fall months (Greg.); 40 
they are only suited for a rough determination of Ramesside chronology. 41 
By contrast, the correct harvest date is crucial for the hypothetical core- 
gency of Thutmose III with Amenhotep II. 42 

33 Willcocks (n. 19), 38 (304). 

34 P. Smither, JEA 27 (1941), 74-76. — R. Krauss, "Detailfragen der altagyptischen 
Getreidewirtschaft", in: Form und Mass. Fsjur Gerhard Fecht, J. Osing & G. Dreyer (eds.), 
(Wiesbaden, 1987), 268-269. 

35 Cf. J. J. Janssen, BiOr 49 (1986), 356, and W. Helck, LA II, 152-153. 

36 Cf. for example, Beckerath, Chronologie, 53. 

37 R. Krauss, "Marz, April und Mai als durchschnittliche Monate der Getreideernte 
im antiken und neuzeitlichen Agypten" DE 27 (1993), 27-34 

38 For threshing dates as late as August/September (Greg.), see Krauss (n. 37). 

39 H. Felber, "Die Daten in den demotischen Ackerpachtvertragen der Ptolemaerzeit 
und das landwirtschafdiche Jahr", APF, Beiheft 3 (1997), 281-289. 

40 For example, pTurin 1895 + 2006 lists collection dates between May 29 and 
September 4 (Greg.); cf. A. H. Gardiner, JEA 27 (1941), 22-37. 

41 Beckerath, Chronologie, 53. 

42 Cf. above Hornung, Chapter II. 8, and below Krauss, Chapter III. 8. 


Wine Delivery Dates 

Since the Middle Ages and down to the present, grapes ripen in Egypt 
towards the end of June; 43 vintaging and wine-making occurred in 
August (Greg.). 44 If the same conditions prevailed in the NK, the sealing 
of the wine jars would have taken place after fermentation towards the 
end of August/beginning of September (Greg.). 45 Jar labels mention the 
regnal year of the sealing; the month is very seldom cited. 46 The labels 
name the chief vintner and the wine estate, but not the ruling king. 
Great quantities of dated wine jars were excavated at Malqata 4 ' and 
Amarna, at the Ramesseum, 48 and in Deir el-Medina. 49 At Amarna wine 
jar labels of 1 + 16 successive vintages were found. The labels document 
the later part of Akhenaten's reign, and the reigns of two successors. 50 
A wine jar from the funerary temple of Amenhotep II bears the date 
"regnal year 26". 51 If the year is ascribed to Amenhotep II, then the 
sealing of the jar took place in ca. II Akhet, 52 10 months after the 
beginning of year regnal 26. 53 It may be surmised that the reign of 25 
y + 10 m that Manetho (Flavius Josephus) ascribes to Mephramuthoses, 
the precursor of Thmosis < Thutmose IV, belongs to Amenhotep II. 

Harvesting of Flax 

In the tomb of the nomarch Djehutinakht (Bersheh 1), a scene of har- 
vesting flax is dated to IV Akhet 23. )4 The nomarch is datable to 
around year 31 of Senwosret I. 55 Meyer 56 and then Borchardt 5 ' relied 

43 Krauss, SAK 23 (1996), 237 n. 67. 

44 C. Pellat, Cinq calendriers egyptiens (Cairo, 1986), 247, s.v. vin. 

43 R. Krauss, MDOG 129 (1997), 227; idem, SAK 23 (1996), 238-239; with addi- 
tional literature. 

46 Cf. Hornung, Untersuchungen, 78 n. 51. 

47 W. C. Hayes, JNES 10 (1951), 41-56. 

48 W. Spiegelberg, %AS 58 (1923), 25-36; Beckerath, Chronologic NR, 40. 

49 Beckerath, Chronologic NR, 40. 

50 See above Hornung, Chapter II. 8. 

51 Urk. IV, 1365; Beckerath, Chronologic NR, 94; idem, Chronologic, 109. 

32 II Akhet 1 = September 9 Jul. = August 26 Greg., in 1400 BC. 
See above Hornung, Chapter II. 8. 
PMW, 177. 

33 H. G. Fischer, LA II (1977), 414. 
Meyer, Nachtrdge, 189-20. 
Borchardt, Mittel, 89-90. 


on information in the Description de I'Egypte that flax harvesting took 
place in the region of Asyut at the beginning of April and lasted 8 to 
10 days. But rather than this isolated date, the general date for flax 
harvesting in Egypt should be used, viz. March to April. 58 Furthermore, 
the date is not necessarily contemporaneous with Djehutinakht, for the 
scene with its date could have been copied from another earlier tomb. 
IV akhet 23 corresponds to March 15 Greg. = March 31 Jul. in 
1893/1890 BC and to April 15 Greg. = May 2 Jul. in 2021/2018 BC. 
Harvesting of flax as early as March 15 Greg, or later is compatible 
with the low chronology (Senwosret I reigned 1920-1875 BC), but also 
with a slightly higher one. 

Seasonal Finds in the Tomb of Tut'ankhamun 

Provided the times of flowering and ripening are known and if the 
material did not come from storage, the season when plant mate- 
rial was deposited in a tomb can be determined in terms of the Egyp- 
tian calendar. By far the most important example is the tomb of 
Tut'ankhamun with its great diversity of plant remains. The flowers 
that were used in the wreaths (Picris asplenoides L., Centaurea depressa, 
Whitania somnifera (L.) and Nymphaea caerulea Sav.) blossom in April; 
the Christthorn fruit (Zizyphus spina Christi), of which great quantities 
were found, also ripens in April. The absence of oil seed and fruit 
common in other tombs, which was harvested in the summer months 
(Balanites Aegyptiaca, Ricinus communis; Cyperus esculentus, Punica 
granatum and Ficus carica), indicates that the burial took place when 
these fruits were already used up. Thus it is to be concluded that the 
burial took place in April (Greg.). 59 Provided the burial followed a reg- 
ular mummification process of 70 days, the king would have died 
between January 2 1 (Greg.) and February 20 (Greg.). If he died in or 
ca. 1322 BC the interval corresponds to III peret 18/IV peret 18 which 
would also subsume the accession date of Aya, otherwise not known. 60 

58 R. Krauss, BSEG 15 (1991), 79; with additional literature. 

59 P. E. Newberry, in: H. Carter, The Tomb of Tutankhamen II (London, 1927), 189-196; 
R. Germer, Die Pflanzenmaterialien aus dem Grab des Tutanchamun (Hildesheim: HAB 28, 
1989), 4-26; R. Krauss, "Nochmals die Bestattungszeit Tutanchamuns", SAK23 (1996), 

60 Cf. above Hornung, Chapter II. 8. 


Tut'ankhamun's death in January/February raises an obstacle to his 
identification with Nipkhururia, who apparently died at the end of sum- 
mer or early in the fall. 61 

Solar and Lunar Eclipses 

There are no reports of solar and lunar eclipses from pharaonic Egypt. 
Thus ancient Egyptian data are absent in modern astronomical analy- 
sis of recorded eclipses. 62 A text from year 15 of Takelot II was thought 
to refer to a lunar eclipse. If the verb in question is understood as 
sdm.f, then the translation is "the sky did not devour the moon", 63 if 
the form is interpreted as n sdmt.f, then the translation is "before the 
sky swallowed the moon". 64 Regardless, it was already known in the 
19th century that there was no lunar eclipse which would fit the stan- 
dard chronology of the period; 6 ' it seems possible that the text refers 
to a delayed appearance of the new crescent. 66 

A solar or lunar eclipse supposedly occurred when Psammetichus I 
died. 67 The source is the Demotic papyrus Berlin 13588, written in late 
Ptolemaic or early Roman times. 68 The papyrus relates how a priest 
heard that the sky swallowed the disk (jtn) when Psammetichus I died; 
later the priest copied the "Book of Breathing" onto the mummy wrap- 
pings of Psammetichus I. The reported time of the supposed eclipse 
does not suit the solar eclipse of September 30, 610 BC; instead, it 
would fit the lunar eclipse of March 22, 610 BC. The context of the 
eclipse is fictitious, insofar as it would date the "Book of Breathing"- 
a creation of the Ptolemaic period — to the time immediately after the 

61 Cf. above Klinger, Chapter II. 13. 

1.2 Stephenson, Eclipses, 58-59, and J. M. Steele, Observations and Predictions of Eclipse 
Times by Early Astronomers, (Dordrecht, 2000), 6. 

1.3 Caminos, Chronicle, 88-89. 

64 K. Jansen-Winkeln, SAK 21 (1994), 127. 

65 Cf. Kitchen, TIP, 181-182; Beckerath, Chronologie, 41-42. 

66 R. Krauss, "Die Bubastiden-Finsternis im Licht von 150 Jahren Forschungs- 
geschichte", MDAIK 63 (2007), in press. 

67 O. Neugebauer, The Exact Sciences in Antiquity' 1 (Providence, 1957) 95 n. 2; E. Hornung, 
ZAS 92 (1966), 38-39; M. Smith, OLP 22 (1991), 101-109. 

1,8 W. Erichsen, "Eine neue demotische Erzahlung", Akademie der Wissenschaften und 
der Literatur Mainz. Abhandlungen der geistes- und sozialwissenschaftlichen Klasse 1956, 
Nr. 2, 49-81. 


death of Psammetichus I. An eclipse that is reported in a fictitious tale 
cannot be deemed historical. 69 

Eclipses that were recorded in the Near East such as the solar eclipse 
of 15 June 763 BC (see above Klinger, p. 308) are of indirect impor- 
tance to Egyptian chronology. The solar omen of year 1 of Murshili 
II would be important for the chronology of the Amarna period and 
its aftermath, if it were indeed an eclipse. The text says that the sun 
god "gave an omen"; the kind of heavenly phenomenon (eclipse?) is 

Miscellaneous Astronomical Phenomena 

The observations of the planets Mercury (Seth) and Venus (Horus, Eye 
of Horus) that are reflected in the Cairo "Calendar of lucky and unlucky 
Days" are datable to 1298/97 BC, whereas the manuscripts are at least 
80 years younger. 71 

There are, for example, no Egyptian reports about periodic comets, 
singular Novae or occultations. Whether the "star miracle of Thutmose 
III" reflects an actual event or phantasy is moot. 72 

J. Herschel dated the Giza pyramids on the assumption that the ori- 
entation of the ascending corridors was determined by the position of 
the "polar star" at the time when the pyramids were laid out.' 3 Herschel 
subsequently retracted his proposal and accepted the explanation that 
the angle of the corridors was chosen to facilitate their filling with large 
blocks after the burial of the pyramid's owner. 74 But this was over- 
looked by later astronomers and Egyptologists who criticized his orig- 
inal idea.' 5 

69 Krauss (n. 66). 

70 Ph. van den Hout, Purity of Kingship (Leiden, 1998) 42ff.; D. Schwemer, in: Die 
Hethiter und ihr Reich. Das Volk der 1000 Cotter. Exhibition catalogue (Stuttgart, 2002), 
140-145, at 144. 

71 R. Krauss, "The Eye of Horus and the Planet Venus: Astronomical and Mythological 
References", in: Under One Sky, J. M. Steele & A. Imhausen, eds., (Miinster: AOAT 
297, 2002), 193-208. 

72 Urk. IV 1238; cf. D. Meeks, iA IV, 117-118. 

' :i J. Herschel, Outlines of Astrononomy 6 (London, 1859), 205-207. 

74 M. Brack, Journal of the British Astronomical Association 105.4 (1995), 161-164. 

75 E.-M. Antoniadi, L'astronomie egyptienne depuis les temps les plus recules jusqu'a la fin de 
I 'epoque Alexandrine (Paris, 1934), 146; J.-Ph. Lauer, Le probleme des pyramides d'Egypte (Paris, 
1952), 187-188. 


According to a hypotheses proposed by K. Spence, the pyramids of 
Dyns. 4 and 5 were oriented towards north utilizing the "simultaneous 
transit method", i.e. projecting the chord joining two circumpolar stars 
on opposite sides of the celestial north pole at the moment of their 
respective upper and lower culmination. 76 In the course of decades and 
due to precession the method would yield first values east (west) of 
north, then true north, later values west (east) of north. If the method 
was applied in the 26th and 25th century BC to the stars Mizar (C, UMa) 
and Kochab ((3 UMi), then the resulting pattern seems to fit known 
pyramid alignments." But the orientation of Djedefre"s pyramid which 
became known in 2001, 7!i does not fit into the sequence, and thus the 
hypothesis cannot be correct. 79 

Moreover, Spence presumes that it was the west sides of the pyra- 
mids which were astronomically aligned; in fact, the corridors on or 
parallel to the central axis are decidedly closer to true north than the 
west sides and it is likely that they were the features to be astronom- 
ically aligned. However, the orientation of the corridors does not adhere 
to a scheme that could have been related to precession and thus the 
pyramids cannot be dated by this means. 80 

Finally, it should be emphasized that the angles of the air shafts in 
the "Great Pyramid" (pyramid of Cheops) are useless for dating the 
structure astronomically. 81 Another senseless exercise involves using 
the presence or omission of certain planets in the representation of the 
astronomical ceiling in the tomb of Senenmut for dating purposes. 82 

76 K. Spence, Orientation in Ancient Egyptian Royal Architecture (unpubl. Diss., Cambridge, 
1997); idem, Mature 408 (2000), 320-324. 

77 Cf. J. A. Belmonte, "On the orientation of the Old Kingdom pyramids", Archaeo- 
astronomy 26 {JHA XXXII) (2001), S1-S20. 

78 E. Aubourg & C. Higy, "Determination de Porientation de la pyramide de 
Redjedef", BIFAO 101 (2001), 457-459. 

79 Cf. also A.-A. Maravelia, "L'horizon astral de Kheops. Archeoastronomie, egyp- 
tologie . . . et quelques scenarios de science-fiction", Tozai 5 (2000), 11-37. 

80 R. Krauss, ",:Las ilusiones perdidas? Recientes intentos en Arqueoastronomia en 
Egipto", BAEDE 16 (2006), in press. 

81 See R. Krauss, "Los canales en la piramide de Kheops", BAEDE 13 (2003), 
55-66, rejecting Briick (n. 74). 

82 As C. Leitz, Studien, 35ff. does; accepted by Beckerath, Chronologie NR, 94 
n. 581; refuted by R. Krauss, GM 146 (1995), 61-70.— O.v. Spaeth, Centaurus 42 (2000), 
159-179; countered by C. Leitz, Centaurus 44 (2002), 140-142. 


Juan Antonio Belmonte 

Astronomy can offer a third dating option, over and above the analy- 
sis of lunar and Sothic dates, viz. the possibility of dating monuments 
using solar, lunar or stellar alignments depending upon the variation 
of stellar coordinates due to precession or the variation of ecliptic obliq- 
uity. For such analysis, astronomy on the horizon is the most relevant 
tool. The pioneering work in this area is Norman Lockyer's The Dawn 
of Astronomy, 1 considered today by some archaeoastronomers as the first 
"serious" book in their discipline. The author made ample use of pre- 
cession to date Egyptian temples to support the long chronology which 
was accepted in his day (Dyn. 1, ca. 5000 BC). When Egyptologists 
discarded such chronologies any possibility of their using archaeoas- 
tronomy as a chronological tool disappeared with it. In the 1970s Gerald 
Hawkins 2 reopened the discussion; although the topic was promoted by 
Edwin Krupp, 3 there was no noticeable response from Egyptologists. 

Astronomical alignments are either directed towards the horizon or 
towards lower zenithal distances, including alignments towards zenith 
pass. The latter has seldom figured in archaeoastronomical studies, 
although it has the advantage of offering fewer problems. But its exis- 
tence is difficult to demonstrate, except in the case of zenith pass for 
which Mesoamerica furnishes several good examples, but Egypt none. 4 

1 J. N. Lockyer, The Dawn of Astronomy (London, 1894). 

2 See e.g. G. S. Hawkins, Beyond Stonehenge (New York, 1973); idem, "Astroarchaeology: 
The Unwritten Evidence", in: A. Aveni, ed., Archaeo-astronomy in Precolombian America 
(Austin, 1975), 131-162. 

3 E. C. Krupp, In search of the Ancient Astronomers (New York, 1977), 208—219; idem, 
"Egyptian Astronomy: Temples, Traditions, Tombs", in: Archaeoastronomy and the Roots 
of Science. AAAS Sym. 71 (Boulder, 1984); idem, Beyond the Blue Horizon (Oxford, 1991). 

4 For Mesoamerica, see A. F. Aveni, Skywatchers of Ancient Mexico 2 (Austin, 1990); for 
a hypothetical use of zenith passage in Egypt, see J. A. Belmonte, "Some open ques- 
tions on the Egyptian calendar: an astronomer's view", Trabajos de Egiptologia 2 (2003), 


The controversial theories relying on simultaneous star transit to explain 
pyramid alignments for dating purposes exemplifies the problem. 5 Part 
of the "Cosmology of Nut" system, present in the tomb of Ramesses 
IV and the Osireon of Abydos, and the Ramesside clock devices in 
the tombs of Ramesses VI, VII and IX probably also related to stars 
in positions far from the horizon. 6 

Astronomy dealing with phenomena near the horizon includes heli- 
acal risings (and settings) which are basic for the understanding of 
decanal star clocks.' However, in what follows I shall concentrate on 
possible stellar or (luni)solar alignments on or near the horizon as a 
means of dating pharaonic monuments. 

In Fig. III. 5.1 there are three problems, two of atmospheric origin 
(refraction and extinction), and one of a topographic nature (rough 
horizon). In the case of (luni)solar observations, the size of the solar 
disc presents a fourth problem. The azimuth a of the rising or setting 
of a celestial object can be calculated in principle by using simple spher- 
ical trigonometry. Consequently, if a building were oriented with ref- 
erence to a certain celestial body, 8 it might be possible to calculate the 
date of the building's foundation. 

Simple spherical trigonometric calculations would apply only to a 
planet without atmosphere and with a flat surface. In reality, a celes- 
tial body is never seen rising or setting at a; Figure 1 illustrates the 
following actual possiblities: 

a': The setting of the star, if only refraction is taken into account, 
e.g., at a flat desert horizon. This also varies because refraction is espe- 
cially dependent on atmospheric conditions such as humidity, temperature, 

' K. Spence, "Ancient Egyptian Chronology and the Astronomical Orientation of 
Pyramids", Nature 408 (2000), 320-324.— Challenged by A. A. Maravelia, "L'horizon 
astral de Kheops", Tozai 5 (2000), 11-37, and J. A. Belmonte, "On the orientation of 
the Old Kingdom pyramids", Archaeoastronomy Supplement, JHA 26 (2001), S1-S20. 

6 See C. Leitz, AlUigyptische Sternuhren (Louvain: OLA 62, 1995); J. A. Belmonte, "The 
Ramesside star clocks and the ancient Egyptian constellations", in: Calendars, symbols 
and orientations: legacies of astronomy in culture, Proceedings of the SEAC 9th Annual Meeting 
Uppsala Astronomical Observatory Report 59 (2003), 57-65. 

7 On the decans see J. A. Belmonte, "The decans and the ancient Egyptian sky- 
lore: an astronomer's approach", Proceedings of the INSAP III Meeting; Memorie delta Societa 
Astronomica Italiana 63 (2001), special vol. I, 43-57. 

8 See e.g., Leitz' controversial Studien, and J. A. Belmonte, JHA 26 (2001), S1-S20. 
See also the only slightiy outdated study of Z. Zaba, L'orientation astronomique dans I'an- 
cienne Egypte, et la precession de I'axe du monde (Prague: Archiv Orientalni Suppl. 2, 1953). 






Figure III. 5.1: Real setting track (black) of a celestial object (here: Venus) under 
extreme atmospheric conditions. (On the basis of a photograph taken on southern 
Tenerife, Canary Islands, Lat. 29".) Compare the theoretical astronomical setting azimuth 
a to the actual azimuth at last visibility of the star at a or the azimuth under ideal 
atmospheric conditions only affected by low refraction and rough horizon a'". The 
theoretical star track is in white; topographical features and size of the sun disk are 
at the same scale; the slightly elliptical shape of the disk is due to refraction. 

and the presence of dust or haze near the horizon, an extremely 
frequent occurrence in Egypt. 

a": The azimuth of the actual last visibility of the star is due not 
only to refraction but particularly to atmospheric extinction. According 
to a rule of thumb, a star becomes visible above the horizon, if its 
angular height is at least equal to its magnitude. Accordingly, under 
the best circumstances only Sirius, Vega, Rigil Kentaurus or Arcturus, 


or the brightest planets, would be visible at 0° altitude. However, this 
too, is very dependent on atmospheric conditions; dust clouds ("calimas") 
or haze can severely affect visibility. In particular, Figure 1 represents 
a setting of Venus in southern Tenerife when the planet had the mag- 
nitude minus 2. However, sighting of Venus was lost when it was still 
more than 2° above the horizon. It was a dusty day, as often occurs 
at Saharan latitudes like the Canary Islands or Egypt. 

a'": Setting azimuth of the star taking into account refraction and 
rough horizon. This would have been the actual setting azimuth of 
Venus on this occasion, provided the atmosphere would have been 
much clearer and more stable on that night. 

a"": Theoretical setting azimuth considering only rough horizon and 
no atmosphere. This value is obtained from standard azimuth (a mra ) and 
height (h mes ), with measurements taken either with a theodolite or a 
tandem when aligning a specific structure. For the sun or the moon, 
one should consider, apart from parallax, the size and shape of the 
disc (the latter dependent also on refraction and extinction) which at 
a rough horizon, can substantially change the position of the last con- 
tact event (or first contact for rising), and, consequently, the alignment 
of a building. 

On this basis I can affirm without reservations that a precision of 
V2 in determination of azimuth is perhaps the best one can expect for 
solar or very bright star observations near the horizon in Egyptian lat- 
itudes. For fainter stars, like those of the Foreleg (mshtyw) or Orion {sih), 
or important asterisms, like the Pleiades (few), the error in azimuth can 
range between one and several degrees of arc. 9 Because of this variation, 
Haack's theory of pyramid orientation was not taken seriously and his 
discovery of the error versus time trend ignored. 10 For the same reason 
Isler 11 and Edwards 12 were forced to abandon horizontal astronomy in 
favor of a cast shadow system or an artificial horizon, respectively. 

9 See the figures in J. A. Belmonte & M. Hoskin, Reftejo del Cosmos: Atlas de 
Arqueoastronomia del Mediterrdneo Antiguo (Madrid, 2002), 25. They are based on R. M. 
Sinclair & A. Sofaer, "A method for determining limits on the accuracy of naked-eye 
locations of astronomical events", in C. L. N. Ruggles, ed., Archaeoastronomy in the 1990s 
(Loughborough, 1993). See also B. E. Schaefer, "Atmospheric extinction effects on stel- 
lar alignments", Archaeoastronomy Supplement J HA 10 (1986), S32-S42. 

10 S. C. Haack, "The Astronomical orientation of the Egyptian Pyramids", Archaeo- 
astronomy Supplement, JHA 7 (1984), SI 19. 

11 M. Isler, "An ancient method of finding and extending direction", JARCE 26 
(1989), 191-206. 

12 I. E. S. Edwards, The Pyramids of Egypt" 1 (Harmondsworth, 1993). 


Does this mean that astronomical alignments are completely useless 
for chronological issues? The answer is yes and no. A recent study by 
Gabolde 13 illustrates the problem for a solar alignment. There is gen- 
eral agreement that Amun's temple at Karnak was aligned towards the 
rising sun on a winter solstice. 14 However, Gabolde's attempt to fix 
the date as the winter solstice following July 27 in 1946 BC, during 
the reign of Senwosret I, for laying the temple axis should be viewed 
with extreme caution. The change of solar declination due to variation 
of ecliptic obliquity is of the order of only 0".46 per year. Therefore 
the change of V2 occurs only after 3900 years, which is approximately 
the time elapsed between Senwosret I and the present. Thus the 
minimum error equals the variation, making dating tenuous. 15 Either 
those who set the axis were extremely good topographers and precise 
technicians working under perfect atmospheric conditions, or I should 
conclude that we are faced with a very lucky situation, one in which 
fine calendrical determination and timing, splendid weather and the 
technical skill of ancient Egyptians conspired to yield an almost per- 
fect alignment. If standard chronology did not support this epoch, I 
would have never been confident about fixing the date of Senwosret 
I's tenth regnal year just using solar alignment. 

For stars, the situation is different. Precession can substantially change 
the coordinates of a certain star and thus its rising and setting azimuths 
or, in general, its position in relation to the local horizon (e.g. angu- 
lar height and moment of culmination). Sirius is exemplary for the case 
of horizontal astronomy. Between 3000 BC and 500 AD, the declina- 
tion of Sirius changed from -22°. 7 to -15°. 8 and its rising azimuth 
changed from II6V2 to IO8V3 for the latitude of Cairo (30°). The ris- 
ing azimuth would change from 115° to 107'/3° for the latitude of 
Philae or Aswan (nearly 24°), where important temples of Isis or Satet 
were located, the divinities who manifest themselves in Sirius. 16 Presuming 

13 L. Gabolde, Le "Grand Chateau d'Amon" de Sesostris I" a Karnak (Paris: MAIBL, N.S. 
17, 1998), 123-137. 

14 For this controversial topic, see M. Shaltout & J. A. Belmonte, "On the orienta- 
tion of ancient Egyptian temples: (1) Upper Egypt and Lower Nubia", JHA 36 (2005), 
273—298, with references; for a different view see R. Krauss, ",jLas ilusiones perdidas? 
Recientes intentos en Arqueoastronomia en Egipto", BAEDE 16 (2006), in press. 

15 According to my personal impression the foundation blocks of the original MK 
temple do not permit a much better precision even with the use of a theodolite. 

16 See R. A. Wells SAK 12 (1985), 255-302, for a study of the orientation of the 
Satet temple at Elephantine with reference to Sirius. 


an error of V2 , we can obtain a precision on average of nearly a quar- 
ter of a millennium or, under excellent atmospheric conditions, per- 
haps a little better. Obviously, the azimuth dating device is not very 
precise, but not useless. If there is textual evidence, as, for instance, in 
Dendera and perhaps in Philae, we might be able to date alignments 
to Sirius within some centuries. 

In the case of fainter stars their coordinate variation and their mag- 
nitudes are decisive, but, in general, the results are less reliable than 
for Sirius. 17 Consequently, for other stars or planets, such us the bright 
stars of the Foreleg, and for other epochs, we can make estimates per- 
haps to the nearest century in the best cases, even though, on some 
occasions, we might be very lucky, as in the example of Karnak. 

Astronomy near the horizon cannot be used as an appropriate, and 
certainly not as a definitive tool by itself for establishing the precise 
parameters of pharaonic chronology. However, the approach is useful 
nevertheless, for it provides insights into the role of astronomy within 
the culture of ancient Egypt, particularly in the religious sphere. 1 " 

17 Except for stars where precession can produce a substantial change in coordinates. 
Alnilan (e Ori), the central star of the s>h constellation, is a good example; its decli- 
nation changed from — 17°.3 in 3000 BC to — 3°.4 in 500 AD. A still better example 
is Arcturus (probably to be identified as a bright star of the constellation mnji), which 
changed from 48".6 to 27". 6 in the same period. In these cases, V2" error in azimuth 
would permit a precision of 110 and 60 years, respectively, for the latitude of Cairo. 

18 See, for example, G. de Young, "Astronomy in Ancient Egypt", in: H. Selin, ed., 
Astronomy across Cultures (Amsterdam, 2000), 475-508; J. A. Belmonte, "Astronomia y 
arquitectura: el papel de los astros en la cultura y el arte del antiguo Egipto", in M. A. 
Molinero & D. Sola, eds., Arte y Sociedad del Antiguo Egipto (Madrid, 2000), 109-136. 




Rolf Krauss 

The Days of the Lunar Month 

The ancient Egyptians observed the phases of the moon; they counted 
and named the days of the lunar month as well. Early on Brugsch 
compiled a list of the names of the lunar days; they are readily acces- 
sible in Parkers's Calendars? 

The earliest attestations for lunar days occur in private and royal 
inscriptions of Dyns. 4 and 5. The Palermo stone preserves the earliest 
royal example: LD 6 = smut, a day of offering at Heliopolis in regnal 
year 6 of Weserkaf. 2 The lunar days psdntyw, >bd, snwt, dnjt and smdt, 
i.e. LD 1, 2, 6, 7 and 15 are attested as days of rituals in the Pyramid 
Texts. 3 Spalinger has collected and analysed the private feast lists of 
all periods. 4 The early lists mention the lunar days ibd and smdt as well 
as s'd, 5 but not psdntyw. As Spalinger notes, "when one descends in 
time from the Old Kingdom to the very last phases of Pharaonic civ- 
ilization, the number of lunar-based feasts diminishes". 6 But lunar days 
are attested throughout Egyptian history and can be utilized for chrono- 
logical analysis, if they are combined with dates of the civil calendar. ' 

1 Parker, Calendars, 11-12; cf. also Belmonte, "Questions", 35. 

2 Wilkinson, Annals, 153-155. — The reading of A. Roccati, La litterature historique sous 
I'Ancien Empire egyptien (Paris, 1982), 43, "dnjt: quarter day" in the entry for year 6 of 
Weserkaf (Cairo fragment) is baseless, cf. Wilkinson, Annals, 219. 

3 For example PT § 657, 716, 794, 1260, 1711, 2056. 

4 Spalinger, Lists, 23-24; 28-29; 33. 

5 Spalinger, Lists, 101-103. 

6 A. J. Spalinger, Studies, VIII, cf. idem, BSEG 19 (1995), 40. 

7 For the non-chronological background to the Egyptian lunar days, see Spalinger, 
"Dating", 383-387. 


Beginning of the Lunar Month 

Historians presumed that the ancient Egyptian lunar month began on 
new crescent day down until 1864 when Brugsch suggested that the 
lunar month started with conjunction. He cited a Ptolemaic text in 
Karnak: "He (Khonsu, the moon-god) is conceived on psdntyw; he is 
born on >bd; he grows old after smelt"/' Ninety years later Parker para- 
phrased the text as follows: The moon-god is conceived in the dark- 
ness of invisibility on the first day of the lunar month, he is born as 
the new crescent on the second day, and he wanes after the day of 
full moon, the 15th day. 9 Parker also cited an earlier, MK text with a 
similar assertion: "I know, O souls of Hermopolis, what is small on 
\3bd] and what is great on [smdt]; it is Thoth." Parker commented: 
"Thoth is, of course, the moon, small on the day of new crescent and 
great on the day of full moon". 10 

Brugsch's contemporaries were less enthusiastic than Parker. 11 Only 
Mahler, and later Sethe, accepted conjunction as the beginning of the 
lunar month. Subsequently others disagreed, arguing that conjunction 
is not observable and thus cannot have marked the beginning of the 
lunar month. 12 Around 1920 Borchardt realized that the Egyptian lunar 
month must have begun with the first day of invisibility after old (or 
last) crescent day, 13 i.e. with an observable event. 14 Shortly thereafter 
Schoch came to the same conclusion independently. 15 Parker argued 
in detail that the Egyptians reckoned the lunar month from the first 
calendar day of the moon's invisibility, coinciding with the day of con- 
junction in ca. 88% of the cases, the day before conjunction in ca. 
10.5% and the day after conjunction in ca. 1.5%. 16 Parker based his 

8 H. Brugsch, Materiaux pour seruir a la reconstruction du calendrier des anciens egyptiens 
(Leipzig, 1864), 58-60; Parker, Calendars, 9, 12. 

9 Parker, Calendars, 9-10. 

10 Cf. also Book of the Dead, title of Spell 135. 

11 Cf. Parker, Calendars, 9. 

12 For example, Meyer, Chronologie, 49-50; D. R. Fotheringham, PSBA 18 (1896), 
101; and E. F. Edgerton, AJSL 53 (1937), 195. 

13 L. Borchardt, OLZ 28 (1925), 620 n. 2; idem, Mittel, 19, 30 n. 10. 

14 Actually two events: the first day of invisibility can only be recognized with cer- 
tainty if the crescent had been observed the day before. 

15 C. Schoch, Die Neumondfeste (Berlin-Steglitz, 1928); Reprint: Aslronomische Abhandlungen. 
Erganzungshefte zu den Astronomischen Nachrichten Bd. 8, Nr. 2 (Kiel, 1931), B11-B13; 
cf. W. Hirschberg, Z^schrifi fur Ethnologic 66 (1933-34), 245. 

16 Parker, Calendars, 9-13. 


argument on the ancient texts cited above and on the correspondances 
of the days of Egyptian terminology for the lunar month with lunar 
phases. 17 There is more circumstantial evidence. For example, in the 
Pyramid Texts psdntyw and ibd occur together as first and second day 
of a statue ritual that is also known from the Nefenrkare c archive. 18 
On psdntyw the royal statues were dressed, and they "appeared" on >bd. 
Apparently, the statues' "appearance" equated metaphorically with the 
appearance of the new crescent on >bd. Thus the earliest known instances 
of psdntyw seem to identify this day as one of invisibility. 

It is astronomically possible that a last crescent is visible in south- 
ern Egypt on a certain day, but not in northern Egypt. 19 In such a 
case, the counting of the lunar days would have been out of step for 
one month. In Egyptian latitudes the new crescent appears in ca. 70% 
of the cases after a single day of invisibility, in ca. 30% of the cases 
after invisibility lasting two days. 20 This circumstance could be under- 
stood to suggest that the lunar month was reckoned from the second 
day of invisibility. But, if the second day of invisibility were counted 
as lunar day 1, how was the first day of invisibility counted? The 
observers did not, and could not, know beforehand whether the moon 
would be invisible for one or two days. This would have been clear 
only on the evening of the day that followed the first calendar day of 
morning invisibility: if the new crescent was observable, the moon had 
been invisible for a single day; if the new crescent did not become 
visible, the moon would be invisible for two days. In other words, the 
theory that the lunar month could have begun on the 2nd day of inivis- 
ibility would mean that for about 36 hours, or IV2 calendar days, 
neither the observers nor the Egyptians who relied on them (for mak- 
ing offerings on the appropriate lunar days) 21 would have known whether 
the current calendar day was to be counted as lunar day **31 or **0 
or 1 or 2. If this impracticable if not to say nonsensical procedure had 

17 Parker, Calendars, 12. 

18 Posener, Archives I, 52—57. — For a chronological analysis of these lunar dates see 
below Chapter III. 8 (end). 

19 For example, the old crescent of December 23 in 1828 BC was visible below 
Coptos, but not in ME and LE. 

20 Parker, Calendars, § 44. 

21 Cf for example, the lunar days 1, 2, 4, 5, and 6 enumerated by King Ahmose 
(Urk. IV 24.4-7) as appropriate for offerings to the dead. 


obtained, then a second day of invisibility that was recorded as apparent 
"lunar day 1" would be equivalent to a positively incorrect lunar date 
(see below, Chapter III. 8). Down to the present no discrepancy of this 
kind has been detected. 

Lunar Calendar: 'Civil-based' Lunar Tear 

Observation of lunar phases, counting of lunar days and lunar-civil 
double dates do not constitute a lunar year that comprises and counts 
successively 12 or occasionally 13 lunar months. The existence of an 
Egyptian lunar year was first suggested by Brugsch; 22 his idea met with 
disapproval. 23 Borchardt developed Brugsch's ideas, but it was Parker 
who argued consistently in favor of two lunar calendars, one referring 
to the heliacal rising of Sirius and one to the first day of the civil cal- 
endar. The ruling principle, as formulated by Parker, was that the 
beginning of the lunar year must not lie before the beginning of the 
civil year. To conform to this rule, a 13th lunar month had to be inter- 
calated from time to time, usually each third year. This lunar calendar 
ran parallel to the civil year, and therefore it may be called a civil- 
based lunar calendar. 

For his interpretation Parker relied on the lunar cycle in pCarlsberg 
9, which dates from 144 AD or later. The cycle consists of 309 lunar 
months or 25 civil calendar years. It begins on I Akhet 1 to which it 
supposedly returns after 25 years. 24 The cycle dates are correct in only 
70% of the cases, as Parker was well aware. 2) The cycle comprises nine 
"great" years of 13 lunar months each and 16 "small" years of 12 
months. Nowadays there are doubts whether it reflects a system that 
was actually used. 2 '' It is equally doubtful whether the same cycle was 

22 Cf. in detail, Depuydt, Calendar, 153-157. 

23 Cf. for example, Ginzel, Handbuch I, 168. 

24 W. Barta, £AS 106 (1979), 1-10, computed a series of 25 year cycles beginning 
in 2388 BC and ending in 155 AD. Because he misunderstood Parker, Barta used 
astronomically incorrect first cycle years; for a correction see idem, CM 94 (1986), 
7-12. Furthermore he did not compute first lunar days, but the respective conjunc- 
tion days, which renders his cycle tables useless. 

2) Parker, Calendars, 25-26. 

2b L. Depuydt, "The Demotic Mathematical Astronomical Papyrus Carlsberg 9 
Reinterpreted", OLA 85 (1998), 1277-1297. 


used for the Macedonian year, which was in effect during in Ptolemaic 
times alongside the Egyptian calendar. 27 

As proof for the existence of the civil-based lunar calendar Parker 
cited double dates that not only counted civil and lunar days, but also 
months differently. 28 The earliest example is the date of an oath taken 
before the moon god Khonsu on a calendric full moon day in year 12 
of Amasis (559 BC). 29 According to Parker, the date is expressed in 
civil and lunar terms as "regnal year 12 of Amasis, (civil month) II 
Shemu 13, being the 15th lunar day of (lunar month) I Shemu". 30 This 
interpretation implies that the months of a lunar year were counted 
from the first LD 1 after (civil) I Akhet 1. 

In 1997 L. Depuydt published a study of the civil-based lunar cal- 
endar, 31 "which subsumes and supercedes previous studies". 32 He con- 
cluded that "there is no doubt about the existence of the civil-based 
lunar calendar, first discovered by Brugsch. The proof is of the best 
kind: astronomical. Civil and lunar double dates are not numerous, 
but they provide unquestionable evidence for the existence of this 
calendar". 33 But as early as 1955 Gardiner had challenged Parker's 
views, rejecting the existence of any lunar calendar. 34 In the late 1980s 
Spalinger began analyzing various aspects of Parker's calendric studies; 
he concluded that no civil-based lunar calendar existed. 33 And recently, 
when J. A. Belmonte scrutinized Depuydt's arguments in favor of the 
civil-based lunar calendar, he also concluded that the data are liable 
to different interpretation. 36 Thus there is no consensus among the 

As Spalinger has pointed out, it is possible to analyse presumed civil- 
lunar double dates without considering the validity of the civil-based 
lunar year itself: 37 ". . . chronographers do not need it, . . . Parker and 

2/ A. Jones, "On the Reconstructed Macedonian and Egyptian Lunar Calendars", 
ZPE 119 (1997), 157-166. 

28 Parker, Calendars, 26. 

29 R. A. Parker, MDAIK 15 (1957), 208-212. 

30 Depuydt, Calendar, 

31 Depuydt, Calendar. 

32 J. P. Allen, The Heqanakht Papyri (New York: PMMAEE 27, 2002), 135 n. 42. 

33 Depuydt, Calendar, 217. 

34 A. H. Gardiner, "The Problem of the Month-names", RdE 10 (1955), 9-31. 

35 Cf. for example, A.J. Spalinger, in: Hommages a J. Leclant (BdE 106/4, 1994), 364 
n. 4. 

36 Belmonte, "Questions", 14-15. 

37 A. J. Spalinger, "Ancient Egyptian Calendars: How many were there?", review 
article of Depuydt, Calendars, JARCE 39 (2002), 241-250, at 250. 


later Egyptologists have never used it . . ." The assertion is correct, in 
general, but the Heqanakht papyri may represent an exception. In his 
letters Hekanakht refers to the months sf-bdt, rkh- c ?, and hnt-hty-prtj. The 
internal chronology of the letters depends on whether these months are 
civil or lunar. For example, Spalinger sees "no indications of a lunar 
calendar operating in this correspondance"; 38 Belmonte likewise asserts 
that the months are not lunar, but "are clearly mentioned in a civil 
calendar context". 39 By contrast, the very use of these month names in 
the Heqanakht papyri suggests to Allen "that farmers also followed the 
lunar calendar — understandably so, since the phases of the moon were 
much easier for them to keep track of than the artificial sequence of 
numbered days in the civil calendar." 40 Allen identifies this presumed 
lunar calendar as the civil-based lunar calendar, without discussing the 
possibility that it might be the Sothis-based lunar calendar. 41 Hekanakht 
would have meant September/ October (Greg.) when he referred to the 
month hnt-hty-prtj as the 10th month of the civil calendar or the civil- 
based lunar calendar; or March/April (Greg.) if hnt-hty-prtj was the 1 0th 
month of the Sothis-based lunar calendar. 42 In the former case Hekanakht 
"postponed the beginning of the household's new salary schedule to 
late September", although it would have been possible to start "two or 
even three lunar months earlier". 43 Perhaps this and other difficulties 
could be resolved by assuming that Hekanakht meant months of the 
Sothis-based lunar calendar, rather than the civil-based lunar calendar. 
This discussion illustrates how the vagueness of the ancient sources, as 
exemplified in the calendric material contained in the Hekanakht papyri, 
are capable of very different interpretations, depending on the bias of 
the scholar. 

38 A. Spalinger, "Calendrical Evidence and Hekanakhte", %AS 123 (1996), 85-96; 
esp. 89. 

39 Belmonte, "Questions", 21. 

40 Allen (n. 32), 135-136. 

41 W. Barta, ZAS 110 (1983), 19. 

42 Based on Allen's assignment of year 8 in the papyri to Senwosret I. 

43 Allen (n. 32), 137 n. 47. 


Kurt Locher 

Of the three sidereally defined periods — the year, the lunar month, and 
the day — , the year is the most stable in terms of modern methods of 
time-keeping such as atomic time based on molecular or electron- 
tilting oscillations. Its slight long-term variation is irrelevant for Egyptian 
chronology, because the artificial 365.0-day civil year, which was kept 
through all epochs, would always yield an unambiguous number of 
days elapsed between any two historical dates of relative chronology. 

This is far from true for the lunar month and the day: The number 
of days elapsed between any pair of the same moon phases many cen- 
turies apart is affected by two long-term variations. The slowing of the 
earth's rotation over time must be studied with the greatest possible 
care, both empirically, using historical astronomical data gathered from 
non-Egyptian records, and theoretically, by calculating its physical cause 
(the amount of kinetic energy released from rotational momentum and 
transformed into heat by tidal friction). Thanks to Stephenson's recent 
extensive work 1 we now know that the actual number of days elapsed 
from any OK date until today differs by just under one day from what 
it would have been if the velocity of the earth's rotation remained 
constant. 2 

Long-term variation in the orbiting period of the moon has essen- 
tially the same physical cause: the slowing of the earth's rotation means 
a decrease in angular momentum, which the motion of the body exert- 
ing the causing force, i.e. the moon, must compensate by acceleration. 
The total number of lunar months elapsed between any OK date and 
today differs from the number they would have amounted to without 
such an acceleration by roughly one hundredth of a month. 3 

1 Stephenson, Eclipses. 

2 Ibidem, Figs. 1.6, 2.1, 14.1, 14.2. 

3 Ibidem, S 2.6.2. 


Since both differences result from an integration over time of a lin- 
ear effect, each increases quadratically with respect to the time elapsed. 
Moving back in time from today towards the OK, the result reduces, 
e.g., to one fourth if we move back halfway. 

Relevant Peculiarities of the Motion of the Moon 

There are also periodic short-term variations in lunar motion. One of 
them arises from the elliptic shape of its orbit and the dynamical con- 
sequences implied which results in variations in speed. 

Since the sun appears, simply spoken in the ancient way, to move 
through exactly 12 zodiacal signs in the course of a year, and since 
there are 12 (exceptionally 13) new moons within this same period, the 
sign in which a new moon is seen from earth is the one adjacent (to 
the left as seen from northern latitudes) to the sign where the preced- 
ing new moon was seen. 

Thus every month the breadth of roughly one sign is covered twice 
by the moon moving through the zodiac; if the moon happens to be 
in that part of its orbit where its speed is smallest (both absolute and 
apparent-angular), the month will be considerably longer than average. 
Such minimum speed occurs near apogee (goto yqv, "away from the 
earth"), the point on the elliptic orbit most distant from the earth. Since 
Ptolemy noted the effect of this variable velocity on the position of the 
moon (or a planet), the phenomenon has been called "the anomaly" 
{av-o\iakoq, "uneven"). 4 Calculating the anomaly in the case of the 
moon's orbit is a complex procedure, because it is considerably per- 
turbed dynamically by the gravity of a third body (the sun); however, 
this perturbation is negligible in the case of a planetary orbital ellipse. 
The perturbation in the moon's case is the reason why the apogee is 
spatially (more exactly: sidereally) not always on the same side of the 
earth, but moves slowly around it, completing a full cycle in roughly 
9 years, a fact which has been termed apsidal motion {oxf\q, "apse" (of 
the ellipse) since Hipparchus. 

The fact that the beginning of an Egyptian lunar month depended 
on the observation of last visibility of the waning moon at dawn must 
also be taken into account. Unlike all other effects considered above, 

4 Ptolemy, Almagest, § IV 2, Edition Heiberg (Leipzig, 1898), 269. 


the success of such an observation depended not only on the lunar 
coordinate measured along the ecliptic, but also at right angles to it, i.e. 
on the ecliptical latitude. Egypt's geographic position, which is to the 
north of the equator, favours observation of the moon if it is situated 
north of the ecliptic, by contrast to a symmetrical position south of it, 
especially for marginal situations which always occur near the horizon.' 

The orbit of the moon is inclined against the ecliptic by roughly 5°, 
so that the moon crosses it twice every month at points called the 
nodes. Like the apogee, these nodes are not fixed, but they revolve 
around the earth in a cycle of roughly 18 years. 

The position of the moon near a node is crucial for the occurrence 
of an eclipse; both solar and lunar eclipses are relevant for the chronol- 
ogy of most ancient civilizations, but, exceptionally, not for Egypt. 
Oddly enough, there are very few, if indeed any reports of eclipses 
from Egypt. For that reason, eclipse theory need not be included here. 

Ibidem, Heiberg, 270. 


Rolf Krauss 

Computation of Old and New Crescent 

An initial attempt to compute old and new crescent was made by the 
astronomer K. C. Bruhns around 1880. 1 He based his calculations, which 
were unsuccessful, on observations recorded at Athens by the astronomer 
J. Schmidt. 15 years later F. Wislicenus also admitted defeat, asserting: 
"if the sky is clear, but under otherwise differing astronomical condi- 
tions, the first appearance of the crescent can occur 1 to 3 days after 
conjunction". 2 It was K. Fotheringham who first succesfully calculated 
old and new crescent, utilizing Schmidt's observations, in addition to 
others. 3 The basic parameters of Fotheringham's calculation are lunar 
altitude and lunar and solar azimuth. 4 For the old or new crescent to 
be considered visible, it must have a minimal altitude h which is depen- 
dent on the distance A in azimuth of the sun and moon at the moment 
when the centre of the sun is in the mathematical horizon; the posi- 
tion of the crescent is computed geocentrically, i.e. without parallax. 5 
M. Maunder, 6 P. V. Neugebauer and C. Schoch' improved upon the 
minimal altitudes h of Fotheringham. In an earlier version of his astro- 
nomical tables P. V. Neugebauer incorporated crescent visibility criteria 
of Fotheringham and Maunder; 8 in his later Astronomische Chronologie, 

1 A. Mommsen, Chronologie: Untersuchungen fiber das Kalenderwesen der Griechen insonderheit 
der Athener (Leipzig, 1883), 69-80. 

2 W. F. Wislicenus, Astronomische Chronologie: ein Hulfsbuch fur Historiker, Archdologen und 
Astronomen (Leipzig, 1895), 29. 

3 Mommsen (n. 1), 69-73. 

4 K. Fotheringham, "On the Smallest Visible Phase of the Moon", Monthly Notices 
of the Royal Astronomical Society 70 (1910), 527-531. 

5 Contemporary astronomers somtimes overlook the fact that Fotheringham's model 
is geocentric, cf R. Krauss, DE 57 (2003), 53-54. — For a topocentric model see J. A. R. 
Caldwell & C. D. Laney: <>. 

6 M. Maunder, Journal of the British Astronomical Association 21 (1911), 355-362. 

7 C. Schoch, Planetentafeln fur Jedermann (Berlin-Pankow, 1927); idem, in: Neugebauer 
(n. 9), I 79, Tafel E 2 1 . — Use of Schoch's data is now known as "Indian method" fol- 
lowing adoption by the Indian Astronomical Ephemeris in 1966. 

8 P. V. Neugebauer, Hilfstafeln zur Berechnung von Himmelserscheinungen. Tafeln z,ur astronomi- 
schen Chronologie III (Leipzig, 1925), VII, Tafel 14. 


devised for the use of historians and astronomers, he used Schoch's 
criteria. 9 Between the publication of Neugebauer's Tables in 1929 and 
ca. 1990 Egyptologists used them for astronomical computations involv- 
ing solar, lunar, planetary, and stellar data. Neugebauer anticipated 
that his Tables would become obsolete after half of a century. 10 And 
indeed the situation changed in the 1980's when astronomical software 
to be used with computers became available. (All astronomical calcu- 
lations for Chapters III. 8 and 10 were made with the program UraniaStar 
Release 1.1 [M. Pietschnig & W. Vollman, Vienna, 1995]. The program 
was developed under the supervision of the astronomer Hermann Mucke. 
For the program's reliability, see M. G. Firneis & M. Rode-Paunzen, 
"Progress-Report on Egyptian Astrochronology", in: Bietak, ed., SCIEM 
Haindorf 2001, 48). 

Especially for lunar positions, modern astronomical computation yields 
results different from Neugebauer's Tables. One reason is that research 
has significantly changed the values for At (delta t), the difference between 
Universal Time and Terrestrial Time that results from the slowing of 
the earth's rotation. 11 Regardless, in many cases the ancient date of 
observability of old or new crescent remains the same, whether calcu- 
lated with the outdated parameters of Neugebauer or according to the 
most recent ones. 12 

In the late 1980's B. Schaefer, following an earlier attempt by 
F. Bruin, 13 developed a model for reckoning old and new lunar cres- 
cents taking into consideration: (1) the physiology of the human eye, 
(2) the brightness of the twilight sky, (3) the surface brightness of the 
moon, (4) the extinction in the atmosphere, and (5) the local conditions. 
According to Schaefer, there is a general shift from clear skies in 
winter to hazy skies in summer, and thus in the northern hemisphere 
the minimal altitude is lower in winter and higher in summer. This 
rule does not seem to apply to all regions, but it is valid for Egypt. 14 
Thus Schaefer's model represents an improvement over the visibility 

9 P. V. Neugebauer, Astronomische Chronologie I, II (Leipzig & Berlin, 1929). 

10 Neugebauer (n. 9, I), V. 

11 See above Locher, Chapter III. 7. 

12 Krauss (n. 5), 52-53. 

13 F. Bruin, Vistas in Astronomy 21 (1977), 331-358. 

14 M. Shaltout, "Study of the Solar Radiation over Menia", Renewable Energy 23 
(2201), 621-639, at 631—634; and personal communication, January 17, 2004. 



criteria of Schoch and others, as far as they do not take into account 
the seasonal variation of extinction and its daily random variability. 
Table III. 8. 1 lists a selection of Schoch's and Schaefer's minimal cres- 
cent altitudes h, relating to A. In the Table months apply only to 
Schaefer's figures; 13 Schoch's do not change during the year. 

For example, if the distance A in azimuth is 0°, then according to 
Schoch the crescent ought to be visible throughout the year if h > 
10.4°. By contrast, Schaefer defines the minimal altitude as a mean 
value h together with its mean square root error o. If A = 0°, then for 
example, in December h = 10.2° + 0.6°, i.e. for ca. 68% of crescents 
h = 10.2° is the minimal altitude; for 16% it would be 10.2° to 10.8°, 
and for the remaining 16% it would be 10.2° to 9.6°. If the crescent 
is within h ± a, then visibility or invisibility of the crescent depends on 
extinction at the time of observation. Thus h + o is a zone of uncer- 
tainty, because extinction cannot be predicted exactly. If the crescent 
is above h ± a, then visibility is to be expected; if the crescent is below 
h + a, it ought to be invisible, although an outlier is a rare possibility. 

Table III. 8.1 





h (A = 0°) 

h (A = 0°) 

h (A = 10°) 

h (A = 10°) 


10.2° ± 0.6° 

8.8° ± 0.8° 

March /Sept 

11.0° ± 0.8° 


9.5° ± 0.9° 



11.6° ± 0.7° 

10.1° + 1.1° 

h (A = 15°) 

h (A = 15°) 

h (A = 20°) 

h (A = 20°) 


7.5° ± 0.7° 

6.6° ± 0.7° 

March /Sept 

8.5° ± 1.2° 


7.6° ± 0.9° 



9.0° ± 0.8° 

7.9° ± 0.8° 

1 ' I owe specific numerical values for h ± a and the permission to use them in pub- 
lications to a generous personal communication from Schaefer in November 1999; see 
also R. Krauss, "An Egyptian Chronology from Dynasties XIII to XXV", in: Bietak, 
SCIEM Vienna 2003, n. 25; in press. 


Negative and Positive Errors in Lunar Observation 

Provided the weather was fine, old crescent ought to have been always 
visible at dawn of a LD 29. The critical day was LD 30. If the cres- 
cent was still visible at dawn of a LD 30, then the current lunar month 
had 30 days and the following lunar month began on the next calen- 
dar day. If the crescent was not visible at dawn of LD 30, then a new 
lunar month had begun and the expired lunar month had 29 days. 
The new crescent appears on the second or third calendar day after 
morning invisibility; an observer could determine the first day of invis- 
ibility with certainty only when he had been able to see the crescent 
on the previous morning. According to Parker, mistakes could occur 
on either day: "An Egyptian lunar date as given in the civil calendar 
and as calculated by modern tables may lack agreement by a day . . . due 
to faulty observation or any other reason". 16 Parker left open whether 
there is a 50% chance that a lunar date is correct, or whether it is 
more probable that a lunar date is correct than wrong. Nevertheless, 
it is clear that Egyptian lunar dates exhibit random qualities, and there- 
fore analysis of them mandates taking into consideration the mathe- 
matical probability of correct and mistaken lunar dates. 

By definition, an ancient lunar date is "correct" if it is confirmed by 
modern astronomical computation. A Egyptian lunar date is negatively 
(or positively) incorrect if the respective lunar phase is one day too 
early (or too late). A negative error results when the observer does not 
see an old crescent that is visible under favorable meteorological con- 
ditions somewhere else. A negatively incorrect lunar date implies the 
beginning of an Egyptian lunar month one day too early. A positive 
error implies a delay of the beginning of an Egyptian lunar month by 
one day, i.e. the month would begin one day too late. A positive error 
should not result from direct observation, because a crescent which is 
not present cannot be seen. A positive error may occur on a cloudy 
day when the observer guesses that the crescent is present above the 
clouds. Late or early lunar dates are evidently not symmetrical errors; 
therefore the likelyhood of each occurring should not be the same. 

Parker, Calendars, 211; idem, JNES 16 (1957), 39-40. 


Error Quotas in Lunar Dates 

The three possible sources of error in Egyptian old crescent observa- 
tion are: a) the observer himself, who might falsify data; b) observa- 
tional difficulties when the crescent is in the zone of uncertainty; c) 
observational difficulties resulting from unfavourable weather conditions, 
i.e. haziness or overcast skies due to clouds or a sandstorm. Presumably 
the accuracy of Egyptian old crescent observation was the same as 
Babylonian new crescent observation. In both cultures observations were 
made under similar meteorological conditions by experienced observers. 17 
C. Schoch 1 " and P. H. Huber 19 investigated the accuracy of Babylonian 
lunar observation. They concluded that at least 85% of the recorded 
lunar dates were correct new crescent observations. Recent analysis sug- 
gests that Schoch used not only observed crescents, but predicted ones 
as well, whereas Huber's data consist of observed crescents only. 20 

F. R. Stephenson's team analysed the circumstances of visibility for 
209 observed Babylonian new crescents 21 published by Hunger. 22 When 
topocentric lunar positions were plotted against different versions of the 
geocentric visibility lines of Schoch, 8 of 209 crescents were below the 
visibility line. If geocentric lunar positions are plotted, only 2 crescents 
are to be found just below Schaefer's visibility line, which if slightly 
adjusted, includes them also. 

According to Schaefer's criteria, ca. 10% of the 209 new crescents 
were sighted one day too late, presumably because of high extinction 
or some other reason. Thus the relative frequency of negative errors 
in these 209 Babylonian crescent observations is ca. 10%. The 95% 
confidence interval for the true probability of negative errors in a sam- 
ple of 209 new crescents is 6% to 14%. Correspondingly it can be 
expected that 86% to 94% of Egyptian old crescent observations were 
correct, the quota of negative errors stemming from extinction being 
6% to 14%. 

17 For the imy-wnwt, the astronomer of Egypt, seeJ.-L. Fissolo, Egypte, Afrique et Orient 
21 (2001), 15-20. 

18 C. Schoch, in: S. Langdon & J. K. Fotheringham, eds., The Venus-Tablets of 
Ammizaduga (Oxford, 1928), 96-97. 

19 P. Huber, Astronomical Dating of Babylon I and Ur III (Malibu: MJNE. Occasional 
Papers 'A, 1982), 25ff. 

20 L. Fatoohi, F. R. Stephenson & S. Al-Dargazelli, "The Babylonian First Visibility 
of the Lunar Crescent: Data and Criterion", JHA 30 (1999), 51-72 at 64. 

21 Cf. n. 20. 

22 Sachs & Hunger, Diaries. 


This result is supported by a computation (sic) of about 150 old cres- 
cents for the years between 2001 and 2013 AD at the latitude of 
Illahun, yielding about 84% crescents (95% confidence interval: 78% 
to 90%) that should (have) be(en) visible without difficulty; about 16% 
of the crescents (95% confidence intervall: 10% to 22%) would (have) 
be(en) situated within the zone of uncertainty. 23 Presuming that high 
and low extinction are equally possible, it is to be expected that in 
84% + 8% = 92% of the cases (95% confidence intervall: 89% to 
94%) the sightings would (have) be(en) correct, and in 8% (95% 
confidence intervall: 5% to 11%) negatively incorrect. Thus it may be 
expected that under favourable weather conditions Babylonian and 
Egyptian crescent observations yielded ca. 90% correct lunar dates and 
ca. 10% negatively incorrect ones. 

In the 1990's Doggett and Schaefer organised "moon watches" to 
establish the Lunar Date Line for specific months. Their evidence for 
negative mistakes is indirect: 24 "Of 520 negative reports, 5 were made 
by observers who missed an easy sighting on the following night. We 
suspect the rate of negative errors is greater (and probably much greater) 
than 1%". The evidence for positive mistakes is direct: "Of the 20 
observers in the northeast [of North America where the moon should 
not have become visible], 3 reported sighting the Moon. In all three 
cases, the reported time of sighting, orientation of the horns, and direc- 
tion of the Moon were grossly in error. The large errors in reported 
details confirm that these three observations were positive errors. From 
our small sample [from Moonwatch 5], the positive error rate is 15%". 
But, the three reports might just as well be spurious; in fact, no mod- 
ern series of observations provides a reliable basis for extrapolating the 
quality of professional ancient Egyptian (and Babylonian) old (and new) 
crescent observations. 

23 Krauss (n. 15). 

24 L. Doggett & B. Schaefer, "Lunar Crescent Visibility", Icarus 107 (1994), 388-403, 
esp. 402; see also B. Schaefer, "Lunar Crescent Visibility", Quarterly Journal of the Royal 
Astronomical Society 37 (1996), 759-768, at 760. 

25 Climates of Africa, J. F. Griffiths, ed., (Amsterdam: World Survey of Climatology 10, 
1972), 84-85, 126-128. 


Mistakenly Conjectured Lunar Dates 

If the sky is overcast, visibility or invisibility of the moon has to be 
conjectured. There are statistics available for the average monthly cloudi- 
ness at Egyptian sites. 2,1 The mean cloudiness decreases between the 
Delta and UE. 26 Presuming the climate in Egypt has not changed since 
the end of the OK, 27 modern regional conditions are applicable to the 
MK. Around Illahun the yearly mean cloudiness amounts to 20%, i.e. 
during the year the probability that clouds obscure old crescent amounts 
to p = 0.2, complemented by the probability of q — 0.8 that old cres- 
cent is visible. Under these circumstances it is to expected that over 
the years, at most ca. 20% of the old crescents went unobserved. 28 

The proportion between lunar months of 29 and 30 days is 47:53; 
the total of 20% old crescents on overcast days divide into 9.4% old 
crescents on a LD 29 and 10.6% on a LD 30. On a overcast LD 29, 
the assumption that the crescent is present is correct, the opposite sup- 
position negatively incorrect. Thus 4.7% of all lunar dates are correct 
and 4.7% negatively incorrect. If it is a LD 30 and the sky is over- 
cast, then the assumption concerning the presence of the crescent is (a) 
correct, and the opposite assumption (b) negatively incorrect. If it is 
not a LD 30 but rather a LD 1 of the next lunar month, then the 
assumption that the crescent is present is (c) positively incorrect and 
the opposite assumption (d) correct. These possibilities result in 5.3% 
correct and 2.65% negatively or positively incorrect lunar dates each. 
Thus the 20% conjectured lunar dates divide into 10% correct dates, 
7.35% negatively incorrect and 2.65% positively incorrect dates. Altogether 
(20% conjectured and 80% observed crescents) there are 82% correct 
dates, 15.35% negatively, and 2.65% positively incorrect dates. The 
percentages of correct and incorrect observations following from 20% 
at Illahun and 12.5% conjectures at Luxor are expressed in Table III. 
8.2 as rounded decimal values. 

26 For example the mean yearly cloudiness in Cairo amounts to 2 octas or 25%, in 
Luxor to 1 octa or 12.5%, (cf. n. 25). 

27 Cf. K. W. Butzer, "Klima", LA III, 456. 

28 The result was obtained by first computing the probabilities that during 62 lunar 
months or ca. 5 years in exacdy 0, 1, 2, 3 ... 62 instances old crescent was hidden 
by clouds. According to standard procedure it follows that the crescent was hidden in 
at most 12 to 14 of altogether 62 instances, corresponding to a mean of ca. 20%. 


Table III. 8.2 



Negatively Incorrect 

Positively Incorrect 





Figs. III. 8.1—3 (below) present the observability of the Illahun lunar 
dates in simplified schemes, according to three different absolute dates 
proposed for year of 1 Amenemhet III. The lunar positions are plot- 
ted against the visibility line of Schaefer. To avoid a three-dimensional 
representation only one visibility line is used and each crescent is plot- 
ted against it, as if it were its own visibility line. Different values for 
At are considered, 29 resulting in different positions of one and the same 
crescent; the positions are represented as a single line corresponding to 
a difference in At of roughly ± 1 h. 

Fig. III. 8. 1 represents the positions of the old crescents when the 
Illahun lunar dates (ILD) are computed as if 1 Amenemhet III = 
1844/43 BC. Fifteen of 21 crescents would have been observable above 
the visibility line. There would have been one negatively incorrect obser- 
vation, since crescent ILD 10 should have been visible in position 10*. 
The crescents ILD 6 and 12 would have been just below the visibility 
line, whereas ILD 13, 14 and 15 would have been far below, alto- 
gether corresponding to five positively incorrect observations. 

Fig. III. 8.2 shows that for 1 Amenemhet III = 1819/18 BC, the 
two crescents ILD 2 and 16 would have been negatively incorrect obser- 
vations. 19 of 2 1 ILD would have been observable above the visibility 
line and none below. 

Fig. III. 8.3 demonstrates that for 1 Amenemhet III = 1794/93 BC 
sixteen crescents would have been visible, two would have been unob- 
servable as positively incorrect and three would have been missed as 
negatively incorrect. 

The situation changes if P. J. Huber is followed with At computed 
according to a formula of Morrison and Stephenson. 50 Then At is 12.6 h, 

29 The lunar positions in Figs. III. 8.1-3 are computed using Spencer Jones's for- 
mula for At, based on an uncertainty of + 1 h. — Cf. H. Spencer Jones, "The Rotation 
of the Earth, and the Secular Accelerations of the Sun, Moon and Planets", Monthly 
Notices of the Royal Astronomical Society 99, (1939), 541-558. 

30 P.J. Huber, Akkadica 119-120 (2000), 173-176; with reference to L. V. Morrison 
& R. L. Stephenson, in: W. Fricke & G. Teleki, Sun and Planetary System (Dordrecht, 
1982), 173-178. 












■ v; ■ 

/» / 

/ distance of sun and moon in azimuth 

i i i i i i i i i i i i i i i i i i i i i i i i i 

Fig. III. 8.1 

s 1 io 




7 / 2 / 

' 3 1 ii 

16*/ > ' ' n 

' • . ^"^-■>^^ ' . visibility line 

distance of sun and moon in azimuth 


5° 10° 15° 20° 

Fig. III. 8.2 





20/ / 






2/ '» { 

15,. -^ U 

7* ' 

. m ^^^^*^^^ ' • t visibility line 


distance of sun and moon in azimuth 


5° 10° 

15° 20° 

Fig. III. 8.3 

instead of Spencer Jones's 10.56 h for ca. 1800 BC. The increased 
value for At results in a decrease of lunar altitude at the moment when 
the centre of the sun is in the horizon. In other words, if the larger 
At is presumed, those crescents in Figs. III. 8.1-3 that are just below 
the visibility line were definitely below it. Of interest is ILD 1 3 in Fig. 
III. 8.2, because it is close to the lower limit of visibility. This crescent 
would have tended to be visible if At < 12.6 h, invisible if At > 12.6 h. 

Partial Repetition of Lunar Dates after 25 Egyptian Years 

The lunar positions represented in Figs. III. 8.1—3 are 25 Egyptian 
years apart; contrary to expectation the positions differ markedly with 
respect to the visibility line. Egyptologists generally assume that lunar 
dates repeat after 25 Egyptian years, 31 because 25 Egyptian years 

31 E.g., K. A. Kitchen, "The Chronology of Ancient Egypt", World Archaeology 23 
(1991), 201-208, at 204: "these moon-risings occur in the ancient calendar every twenty- 
five years." 


correspond to 309 mean (synodic) lunar months of 29.53059 days each, 
the difference being only about an hour: 32 

25 Egyptian years = 25 x 365 days = 9125 days = 309 mean synodic 
months = 309 x 29.53059 days = 9125 days minus 1 hour and 
7 minutes 33 

Actually an observer counts either 29 or 30 full days in a lunar month 
and arrives at the length of the mean synodic month by calculation 
based on the observation of a great number of lunar months of either 
29 or 30 days. 

A lunar date repeats on the same calendar day, if 9125 days com- 
prise 309 lunar months of which 164 are lunar months of 30 days and 
145 are lunar months of 29 days: (164 x 30 days) + (145 x 29 days) = 
4920 days + 4205 days = 9125 days. Because the movement of the 
moon is irregular, there can be 165 lunar months of 30 days and 144 
lunar months of 29 days in a series of 309 lunar months. If so, a lunar 
date does not repeat on the same calendar day after 25 years =9125 
days, but rather after 9126 days. 34 Or if there are 163 lunar months 
of 30 days and 146 lunar months of 29 days in a series of 309 lunar 
months, then a lunar date does not repeat after 25 years on the same 
calendar day, but rather after 9124 days. 

Apparently the irregularity results because the mean synodic move- 
ment (a) comprises (b) the anomalistic and (c) draconitic movement of 
the moon, which do not share a common period of 9125 days = 25 
Egyptian years. The mean anomalistic velocity is not the same after 
25 years, whereas the mean draconitic movement results in a different 
latitude of the moon. 

9125 days = (a) 309 x 29.53059 d + 0.04 d = 

(b) 331 x 27.55455 d + 4.44 d = 

(c) 335 x 27.21222 d + 8.91 d 

On average only about 70% of the dates in a set repeat on the same 
day after a single 25 year shift. 30 For multiples of 25 years, percentages 

32 So for example, Beckerath, Chronologie, 48-49. 

33 Within the siderial and synodic months the moon travels at a mean velocity of 
13.176° per day. Within 25 Egyptian years = 24.982 siderial years, the sun travels in 
the mean 24 x 360° + 353.683°, whereas the moon travels 333 x 360° + 354.272°. 
In 25 Egyptian years the positions of sun and moon have decreased by about 6.317° 
and 5.728° respectively, whereas their original distance has decreased only by about 
0.52°, i.e. the distance that the moon travels in an hour. 

34 For a specific example see Krauss, in: Bietak, SCIEM Haindorf 2001, 190—192. 
33 Krauss, Sothis, 27; similarly Parker, Calendars, 25-26. 



of correct repetitions decrease. Shifts of 2 x 25 and 3 x 25 years yield 
exactly repeated dates in only 50% of the cases. 36 Under these premises 
a large set of Egyptian lunar dates tends to have one solution with a 
maximum of correct dates whereas shifts of + 25 years have less cor- 
rect dates. This is exemplified by 37 alternative solutions for the Illahun 
lunar dates between 2286 and 1387 BC. The calculations use 37 alter- 
natives for 1 Amenemhet III, each separated by 25 years from the next. 
As Fig. III. 8.4 shows, the alternatives differ in the percentages of 
correct and incorrect dates, both negative and positive. There is one 

100% i 

50% " 

0% J 

1843/42 BC- 

1793/92 BC 

-2267 BC 


1818/17 BC 

correct dates 

negatively incorrect dates 

positively incorrect dates 

1368 BC 

Fie. III. 8.4 

36 Krauss, Sothis, 27. 



set with ca. 90% correct dates (1 Amenemhet III = 1819/18 BC), 
whereas the remaining 10% are negatively incorrect. There are seven 
sets with 60% to 80% correct dates and varying percentages of nega- 
tively and positively incorrect dates. The other sets display small per- 
centages of correct dates, whereas their percentages of negatively or 
positively incorrect dates are high. The trend is clearly to more nega- 
tively incorrect dates for shifts backwards and to positively incorrect 
dates for shifts forwards. An increase in the number of correct lunar 
dates after multiples of 150 Egyptian years is to be noted. This results 
from the fact that the synodic, anomalistic, and draconitic months have 
an approximate common period of 150 Egyptian years. Nevertheless, 
instead of yielding a series of various solutions, which are all astro- 
nomically equally possible, there is practically only one astronomically 
workable solution for the Illahun lunar dates. 

The different astronomical possibilities can be evaluated by com- 
puting the respective probabilities. The appropriate tool to deal with 
lunar dates and their three properties is the trinomial formula for prob- 
ability P: 

" ! xplxpixpl 


n = x + y + z;x = correct lunar dates; y = negatively incorrect lunar 
dates; z = positively incorrect lunar dates; pi, P2, and p 3 are the proba- 
bilities of x, y and z (see Table III. 8.2). Table III. 8.3 contains the 
respective probabilities if the ILD are computed for different first years 
of 1 Amenemhet III, including Luft's suggestion that this might be 
1855/54 BC. 37 As Table III. 8.3 implies, only 1 Amenemhet III = 
1819 BC + 25 years is by any means probable (cf. also Figs. III. 8.1—3). 

Table III. 8.3 
1 Amenemhet III correct neg. incorrect pos. incorrect probability 

1869/68 BC 




6.2 x 10-' 

1855/54 (Luft) 



2.7 x 1(T 20 




















Luft, Fixierung, 228; cf. below Krauss, Chapter III. 10. 

408 rolf krauss 

Recorded Lunar Dates from Dyn. 5 to Dyn. 22. 

Dates of a Lunar Feast (Tepi Shemu) in the Bubastide Period™ 

The reign of Psammetichus I began in February 664 BC and the reign 
of Taharqa, 26 years earlier, in 690 BC. According to the Tang-i Var 
inscription, Shebitku ruled from at least 707/706 BC. The highest 
attested date for Shabaka is year 15. Thus 1 Shabaka corresponds to 
722/721 BC at the latest. It is possible, but not assured that Shabaka 
defeated Bocchoris in the former's year 2; 39 thus 6 Bocchoris might be 
723/722 BC = 2 Shabaka or slightly earlier. 

The Bocchoris-Apis was the successor of the Apis that died in 37 
Shoshenq V. According to data concerning the three Apis bulls buried 
between 28 Shoshenq III and 37 Shoshenq V, and presuming that 
Pami's reign ended in a year 7, 95 years elapsed between 1 Shoshenq 
III and 37 Shoshenq V (inclusive). If the Bocchoris-Apis was born very 
soon after the death of its predecessor and had a life span of 26 years 
at most (the maximum life span attested), 40 the upper limit for 1 Shoshenq 
III is 723/722 BC + 26 + 95 = 844/843 BC. This limit would need 
to be adjusted upwards, if Shebitku's reign began before 707/706 BC 
and/or Shabaka occupied the throne longer than 15 years. 

When reckoning the lower limit, it must be borne in mind that 
Shepsesre c Tefnakhte may or may not have ruled in Memphis as pre- 
decessor of Bocchoris for at least 7 full years. Tefnakhte's initial take- 
over of Memphis occurred at the earliest in the course of 38 Shoshenq 
V. In his 20th (?) year Piye drove an apparently non-royal Tefnakhte 
out of Memphis; subsequently the kings Shepsesre c Tefnakhte and 
Bocchoris may have ruled for at least 12 full years (7+5) or more before 
the death of the Bocchoris-Apis. Then, the lower limit for 1 Shoshenq 
III would be ca. 722 BC + 12 + 2 (?) + 95 = 831 BC or 722 BC + 
5 + 2 (?) + 95 — 824 BC, if Shepsesre c Tefnakhte did not rule as king 
in Memphis. 

Aston concluded that the rival kings Takelot II and Petubaste I ruled 
Thebes when Shoshenq III reigned in LE. The synchronism 5 Petubaste 

38 For the TIP cf. above Jansen-Winkeln, Chapter II. 9, and Zibelius-Chen, Chapter 
II. 12. 

39 Krauss (n. 15). 

10 J. Vercoutter, "The Napatan Kings and Apis Worship", Rush 8 (1960), 62-76 at 
64; idem, MDAIK 16 (1958), 339-342. 


I = 12 [Shoshenq III] is attested while [4 Takelot II = 1 Shoshenq 
III] or [5 Takelot II = 1 Shoshenq III] is deducible. 41 If 1 Shoshenq 
III lies between 844 and 824 BC, then 1 Petubaste I falls in 837 to 
817 BC and 1 Takelot II in 848 to 827 BC. The lunar dates of the 
Tepi Shemu feast provide a means for determining exact dates. 42 The 
feast is documented at Karnak from the NK to the Saite Period; 43 
according to MHC 145 Iff., it began on LD 1 and lasted till LD 5. 
Parker computed the latest known example which is dated to 14 
Psammetichus I. 44 Vermis and Krauss followed Parker to interprete 
Bubastide examples of the Tepi Shemu feast as lunar dates. 4,1 According 
to Vernus and Kruchten, the inductions of priests took place during 
the Tepi Shemu feast. 46 Bubastide examples of the *Tepi Shemu feast 
and/or of inductions are: 47 

*(A) 1 1 Takelot II: I Shemu 11 (B) 7 Petubaste I: I Shemu [1] 
(C) 8 Petubaste I: I Shemu 19 *(D) 39 Shoshenq III: I Shemu 26 

The figures in Table III. 8.4 assume that 1 1 Takelot II fell between 
838 and 817 BC. It lists the lunar days which could correspond to the 
explicit Tepi Shemu feast date A; there are six possible years in which 
LD 1 to 5 (± 1 day) corresponded to A (bold type). (Lunar days are 
counted forward as positive from lunar day 1 to 15; starting with the 
last lunar day, whether day 30 or 29, the lunar days are counted back- 
ward as negative down to lunar day 16.) 

Kruchten identified D as a day of the feast of Tepi Shemu. iH The dis- 
tance between A and D amounts either to 32 a + 15 d (if 5 Takelot 

II = 1 Shoshenq III) or to 33 a + 15 d (if 4 Takelot II = 1 Shoshenq 
III). In the latter case the lunar day of D is ca. 1 1 days later than 
that of A, i.e. D is not the date of the Tepi Shemu feast, and neither is 
B or C. In the former case the lunar day of D nearly coincides with 
that of A; thus A and D are days of the Tepi Shemu feast as well 
as the dates B and C. Table III. 8.5 contains the acceptable and the 

41 J. v. Beckerath, "Beitrage zur Geschichte der Libyerzeit", GM 144 (1995), 7-13. 

42 Krauss (n. 15); Krauss, Sothis, 168-177, is outdated. 

43 Schott, Festdaten, 104-105. 

44 Parker, Oracle Papyrus, 7-8. 

15 P. Vernus, "Inscriptions de la Troisieme Periode Intermediaire (I)", BIFAO 75 
(1975), 24; Krauss, Sothis, 168-177. 

46 Vernus (n. 45); Kruchten, Annates, 244 n. 3. 

47 G. Daressy, RecTrav 35 (1913), 130; Kruchten, Annates, 239-240: B-D. 

48 Kruchten, Annales, 80. — Cf. also Spalinger, "Dating", 393. 



Table III. 8.4 

1 1 Takelot II LD 11 Takelot II LD 11 Takelot II LD 11 Takelot II LD 

















































Table III 

. 8.5 

Takelot II A 

B C 


1 Takelot II A 


C D 

845 1 

842 3 
839 5 

5 4 

7 7?8? 
11 10 



834 -1 

831 2 



4 1 

6 3? 

Table III 

. 8.6 







c — 


nearest unacceptable possibilities which result when the lunar day equiv- 
alents of the feast dates of Shoshenq III and Petubaste I are also com- 
puted and correlated with those of Takelot II within the limits for their 
first regnal years. 

There are only two acceptable alternatives: either 1 Takelot II cor- 
responds to 845 or to 834 BC. In 834 BC two of four lunar dates 
would have been negatively incorrect (cc— ), but in 845 BC there is 
only one error of this kind (ccc— ). Table III. 8.6 lists the probabilities 
for the combinations of correct (c) and negatively incorrect (— ) lunar 
dates for Luxor as the place where the observations were made (see 
Table III. 8.2). 

1 Takelot II = 845 BC is definitely preferable, because it is far more 
probable that exactly 1, instead of exactly 2, of 4 lunar dates are too 
early. If Takelot III = 845 BC, the Julian calendar equivalences for 
AD are: 



*(A) 1 1 Takelot II: 
I Shemu 1 1 

(B) 7 Petubaste I: 
I Shemu [1] 

(C) 8 Petubaste I: 
I Shemu 19 

*(D) 39 Shoshenq III: 
I Shemu 26 

LD 30 = November 25, 835 BC 

LD 5 = November 13, 828 BC 

LD 4 = November 11, 827 BC 

LD 1 = December 2, 803 BC 

The figures imply that 1 Shoshenq III = 841 BC and 38 Shoshenq 

V = 743 BC, so that ca. 21 to 20 years elapsed between 38 Shoshenq 

V and 6 Bocchoris. (Kitchen maintains that Shoshenq III ruled after 
Takelot II. 49 Under his premise the Tepi Shemu dates A-D result in a 
single astronomical solution: 1 Takelot II = 856 BC; dead reckoning 
yields ca. 951/953 BC instead of 945 BC for 1 Shoshenq I which is 
not acceptable.) 

The Lunar wrs Date in Year 5 of Shoshenq [IJ i0 

The larger Dakhla stela can be ascribed to Shoshenq I with confidence. 51 
The text mentions a wrs feast on the occasion of a procession of the 
god Seth on IV Peret 25 in 5 [Shoshenq] I.' 2 According to other attes- 
tations, wrs designates the lunar month and/or a lunar day, possibly 
day 1: Demotic papyrus CG Cairo 3080 1; 53 magical papyrus London- 
Leiden (X 22 and XXI 19); 54 Spiegelberg, Mythus vom Sonnenauge, 
VIII 20; M Ny Carlsberg Tebtunis-Papyri;' 6 Demotic Chronicle II, 

49 Kitchen, TIP 3 XXIII-XXV.— Contra Kitchen, cf. G. F. Broekman, GM 205 
(2005), 21-33. 

M For Dyn. 2 1 see above Jansen-Winkeln, Chapter II. 9. 

51 T. L. Sagrillo, The Reign of Shoshenq I of the Egyptian Twenty-second Dynasty (Leuven: 
Dissertation, 2005). — R. Krauss, "Das oro-Datum aus Jahr 5 von Shoshenq [I]", DE 
62 (2005), 43-48. 

52 A. Gardiner, JEA 19 (1933), 19-30. 

>:i Parker, Calendars, § 89—98; cf. also R. Parker, A Vienna Demotic Papyrus on Eclipse 
and Lunar Omina (Providence: BES 2, 1959), 8-9. 

,4 F. LI. Griffith & H. Thompson, The Demotic Magical Papyrus of London and Leiden 
(London, 1904), 79, 139; J. H.Johnson, in: H. D. Betz, ed., The Greek Magical Papyri 
in Translation (Chicago, 1986), 209, 213, 230. 

" J- F. Quack, in: Res severa verum gaudium. Fs fur Karl-Theodor ^auzich, F. Hoffmann 
& H.J. Thissen, eds., (Leuven: SD 6, 2004), 50-51. 

56 J. Osing, Hieratische Papyri aus Tebtunis I. Text (Copenhagen: CNIP 17, 1998), 207-210. 


9; 57 building inscription of Parthenios (Moscow stela).' 8 Thus it is fea- 
sible to search for a match between the date mentioned and a lunar 
day appropriate for a procession. The sum of the highest attested reg- 
nal dates for Osorkon II, Takelot I, Osorkon I, and Shoshenq I, added 
to 841 BC as year 1 of Shoshenq III, yields 934 BC at the latest for 
year 5 of Shoshenq I. Between 950 and 930 BC, only IV Peret 25 - 
December 5 in 939 BC is an acceptable match for the wrs date as 
shown in Table III. 8.7. Consequently 1 Shoshenq I began in November 
943 BC at the latest, and at the earliest in December 944 BC. 

The stela records a judgment on the ownership of a well in Dakhla. 
According to Gardiner's understanding of the text, the mother of the 
claimant is mentioned as the owner in a document dated to year 19 
of a king Psusennes. 39 Because at least 80 years separate 5 Shoshenq 
I and 19 Psusennes I, it is unlikely that the document was written in 
19 Psusennes I, but rather in 19 Psusennes II. 

Three Lunar Dates of the Tepi Shemu Feast in Dyn. 21 

Two of the priestly inductions known from Dyn. 2 1 apparently occurred 
during the lunar Tepi Shemu feast: 60 (a) 2 c Akheperre c setepenre c : I Shemu 
20, induction of a man, (b) whose son was inducted in 1 7 Siamun : I 
Shemu [1]. E. Young assumed that the two inductions would have 
been separated by 20 to 30 years, i.e. a generation. 61 Provided that 
both dates correspond to lunar days 1 to 5, 62 at least 21 y minus 19 
d separate them; other astronomical possibilities are 24, 27 and 30 
years. The distance of 2 1 years is methodologically preferable, because 
young men were inducted when they were 20 years old; 6i furthermore 
a distance of 21 years yields 6 regnal years for c Akheperre c setepenre c , 
as in the Manethonian tradition for "Osochor", successor of Amenemope. 

'" H. Felber, "Die Demotische Chronik", in: Apokalyptik und Agypten, A. Blasius, ed., 
(Leuven: OLA 107, 2002), 76-77. 

58 W. Spiegelberg, £45 66 (1930), 422-443; Borchardt, MM, 39. Parker, Calendars, 18. 

59 Gardiner (n. 52), 28. 

60 Kruchten, Annates, 45-48. — Confirmed by Spalinger, "Dating", 393. 

61 E. Young, JARCE 2 (1963), 99-1 11. 

62 Induction date b cannot be later than LD 3, because the 9th lunar month could 
not begin before IV Peret 28. On the other hand, the distance between induction 
dates a and b implies that b is 3 to 2 lunar days earlier than a. 

63 Kruchten, Annates, 206. 


Table III. 8.7 

Year BC IV Peret 25 Year BC IV Peret 25 Year BC IV Peret 25 


LD 2 





LD 3 





































Provided that 1 Shoshenq I = 943 BC and that Psusennes II ruled at 
least 19 years, induction date b occurred at the earliest in 962 BC. 

Table III. 8.8 contains the LD equivalents for dates a and b. 64 
Between 981 and 962 BC there are three viable alternatives (bold type). 
Only one of them is left, if the oracular text no. 6 of Djehutimose is 
taken into consideration. 65 According to Kruchten, the oracular text 
implies that the last and fifth day of Tepi Shemu fell on I Shemu 10 (or 
9 or 8) in year 3 of either Amenemope, Osorkon, or Siamun. If the 
induction dates (a) and (b) are correctly ascribed to Osorkon and Siamun 
in relative chronology, then the oracular date would correspond to LD 
8—10 in 3 Siamun. It would correspond to LD 15—18 in 3 Amenemope 
if he ruled 9 full years and a fraction of a year, as usually assumed; 
the date would correspond to LD 4-6 if 1 1 years are ascribed to 
Amenemope. By contrast the oracular date coincides with LD 1 to 5 
in year 3 of Osorkon, if induction date (a) corresponds to LD 1 to 5, 
i.e. the oracular date must be attributed to Osorkon. 

In table III. 8.8 only years 970, 973 and 981 BC suit induction date 
(b) in 17 Siamun, and 990, 993 and 1001 BC for induction date (a) 
in 2 Akheperre c (Osorkon). The corresponding distances between 17 
Siamun and 1 Shoshenq I = 943 BC amount to 26, 29 or 37 years. 
If a reign of 1 9 years is supposed for Siamun on the basis of Manetho, 66 
then 24, 27 or 35 years are possible for Psusennes II. But because 
Djehutimose oracle date no. 6 does not work with 992 and 1000 BC 
as year 3 of c Akheperre c (Osorkon), we are left with 989 BC = 3 
c Akheperre c (Osorkon) as the only possibility. 

64 The distance is throughout 21 years minus 19 days. 

65 J.-M. Kruchten, Le grand texte oraculaire de Djehoutymose: intendant du domain d'Amon 
sous le pontifical de Pinedjem II (Brussels: MRE 5, 1986), 237. 

66 Cf. above Jansen-Winkeln, Chapter II. 9. 



Table III. 8.8 

17 Siamun date b 2 Akheperre date a 3 Akheperre oracle date 

962 BC 

LD 28 

982 BC 

LD 1 













































































989 BC LD 5 (6?) 


LD 4 


LD 6 

Ramesside Lunar Dates 1 " 1 

About 85 years elapsed between 1 Smendes and 10 Amenemope. 200 
to 201 years separated the accession of Ramesses II and the latest attes- 
tation of Ramesses XL If these figures are added to 992/991 as year 
1 of 'Akheperre Osorkon, then 1 Ramesses II fell in 1279/1277 BC. 
The ship's log pLeiden I 350 records a LD 1 that coincided with II 
peret 27 in 52 Ramesses II; on that day the ship moored in Piramesses, 
hence the designation Piramesses date. 68 This lunar date can be com- 
bined with dates that refer to the feast-of-the-valley. 69 According to 
MHC 135 the feast began on LD 1 in II Shemu. On that day the cult 
statue of Amun crossed the Nile, went to the temple of Djeser-akhet,'" 
toured Deir el-Bahri in a procession, and spent the night in the funer- 

67 For the NK see above Hornung, Chapter II. 8. 

68 J- J- Jansen, Two Ancient Egyptian Ship's Logs (Leiden: OMRO Suppl. 42, 1961), 
12, 33. Cf. also Beckerath, Chronologie, 51. 

69 For a preliminary analysis under the outdated supposition that the reign of Ramesses 
V extended into a year 5, see Krauss, Sothis, 136-144. 

70 Urk. IV, 929. 


ary temple of the ruling king 71 where Amun received offerings on LD 
1 and 2, according to MHC 159. Four Dyn. 19/20 graffiti from the 
Djeser-akhet temple in Deir el Bahri (DB) attest spending the night 
or receiving offerings in II or III Shemu; 72 the dates imply lunar days 

I or 2. 

DB 3: year 7, II Shemu 28; Amun rests in the funerary temple of 

Twosre 73 
DB 10: year 7, III Shemu 9; Amun rests in the funerary temple of 

Ramesses III' 4 
DB 9: year 6, III Shemu 9; Amun rests in the funerary temple of 

Wesermare setepen ///' 5 
DB 32: year 3, II Shemu 20; Amun rests in the funerary temple of 

Ramesses II 

Combining the Piramesses date with DB 3 and DB 10 allows the 
chronology between Ramesses II and Twosre to be fixed. /b According 
to DB 10 Amun rested in the funerary temple of Ramesses III in a 
year 7, presumably of Ramesses III himself. 7 ' The dead-reckoned min- 
imum distance between DB 3 and DB 10 amounts to either 9 years + 

I I days or 1 a + 1 1 d. The astronomically correct distance between 
DB 3 and DB 10 is 10 a + 11 d = 124 LM - 0.8 d = 3661 days, 
resulting in a correct LD 2 for DB 10, provided that 1 Ramesses II — 
1279 BC. According to oCG 25293 the highest attested dates for Queen 
Twosre are either IV [prt x] and [I] Shemu [y] or IV [Shemu x] and 
[IV] Shemu [y] of regnal year 8; 78 thus the queen ruled either 323 ± 
15 days or 412 ± 15 days subsequent to the date of DB 3. Sethnakhte 
ruled into a year 3, corresponding to a reign of at least 731 days or 
an arithmetical mean of 912 days. There are 2234 days between 
Ramesses Ill's accession and DB 10. These distances add up to 323 
± 15 (412 ± 15) days + 731 to 912 days + 22343 days = 3288 ± 15 
days to 3558 ± 15 days, approximating the distance of 3661 days 
between the lunar dates of DB 3 and DB 10. 

71 Or in the funerary temple of an earlier king, if the ruling king had not yet a 
temple of his own. 

72 Marciniak, Inscriptions. — Peden, Graffiti, 106—107, 120—123. 

73 KRI IV 376. 

74 KRI V 337. 

75 KRI "VI 102; but cf. Peden, Graffiti, 122 n. 395. 

76 For details, see R. Krauss, SAK 24 (1997), 175-177. 

77 Thus also KRI TV, 376. — Cf. table III. 10.10 for the astronomical possibility that 
DB 10 belongs to Ramesses VII. 

78 Beckerath, Chronologie JVS, 74-76. 


If Amenmesses was a usurper in control of Nubia and UE in the 
mid-reign of Sety II, then the interval between the Piramesses date and 
DB 3 amounts to 36 a +121 d = 449 LM + 2 d. This figure corre- 
sponds to the proper interval between a first LD and a LD 1—2 as a 
feast-of-the-valley date (+ 1 day). If, by contrast, the interval between 
the two dates is lengthened by a chronologically independent 4 year 
reign of Amenmesses, then the interval amounts to 40 a + 121 d = 
498 LM + 15 d: then DB 3 would correspond to a LD 15 (full moon), 
but not to a LD 1 or 2 as expected. If the distance is shortened, then 
DB 3 would coincide with a LD 5 or 23. It follows that the reign of 
Sety II subsumes that of Amenmesses. 

Table III. 8.9 presents astronomically possible years for DB 3, 9, 10, 
and 32, provided that 1 Ramesses II = 1279 BC and that the only 
unresolved issue in Ramesside chronology is whether Ramesses' VII 
reign extended into a year 9 and/or Ramesses' VIII reign into a year 
2. Possible years are also given for DB 31, a graffito dated to year 22, 

II Shemu 22 (sic); written "during the feast-of-the-valley". Since neither 
the resting of Amun nor offerings to him are mentioned, it is unclear 
whether the date can be considered to be a LD 1-2. 

Table III. 8.9 demonstrates that DB 9 is attributable to Siptah (sic) 
and DB 32 to Ramesses VI, whereas DB 10 could be a date of Ramesses 

III or VII. The attribution of DB 3 1 to Ramesses XI is problematic, 
because a rock fall might have already destroyed Djeser-akhet during 
his reign. 79 If attributed to Ramesses II, then the date relates to a LD 
3 which might have been a feast day in early Dyn. 19. If attributed 
to Ramesses III, "during the feast-of-the-valley" could relate to all of 
(lunar) II Shemu as the month of the feast. The Julian calendar equiv- 
alents for the lunar days of the graffiti are: 

DB 9: III Shemu 9, Year 6 

[Siptah] : 

LD 2: 

April 21, 

1192 BC 

DB 3: II Shemu 28, Year 7, Twosre: 

LD 2: 

April 10, 


DB 10: III Shemu 9, Year 7, 

[Ramesses III]: 

LD 2: 

April 18, 


DB 32: II Shemu 20, Year 3 

[Ramesses VI]: 

LD 2: 

March 21 

, 1143 

[DB 31: II Schemu 22, Year 22 

[Ramesses XI]: 

LD 2: 

March 8, 


79 Cf. J. Lipinska, JEA 53 (1957), 28-30; G. Pinch, Votive Offerings to Hathor (Oxford, 
1993), 10-11. 


Table III. 8.9 










































DB 3 DB 9 DB 10 DB 32 DB 31 

Ramesses II 


Sety II 



Twosre 2 


Ramesses III 21 2 28 2(i 

Ramesses IV 

Ramesses V 

Ramesses VI 

Ramesses VII 

Ramesses VIII 

Ramesses IX 

Ramesses X 

Ramesses XI 

The DB-graffiti document the feast-of-the-valley in II and III Shemu, 
whereas in MHC 135+159, only II Shemu is mentioned. As Table III. 
8.10 shows, the feast dates fall in LM 11 or 10. The respective LM 
could have been determined by the beginning of the first lunar month 
within the civil year. The feast might have begun on LD 1 in LM 1 1 , 
if the first LM in the civil year began on I Akhet 15 at the latest; if 
after I Akhet 15, then the feast began on LD 1 in LM 10. The ref- 
erence to II Shemu in MHC might reflect the situation in the year 
when the calendar was devised. 

The lunar dates of DB 3, 9, 10 and 32 (DB 31 being not consid- 
ered) are astronomically only compatible with 1 Ramesses II = 1314 
or 1279 BC. In Table III. 8.11 all alternatives between 1304 and 1265 
BC are tested. If 1 Ramesses II were 1290 or 1265 BC, then the DB 
graffiti dates would coincide with LD 3-4 (positive errors), instead of 
LD 1—2. If 1 Ramesses II were 1268 BC, the graffiti dates would cor- 
respond to LD 1-2, but the Piramesses date would be off by +2 days 
(positive error). It follows that only 1304 and 1279 BC are astronom- 
ically possible for 1 Ramesses II. But 1304 BC is excluded on the basis 
of the relative chronology of Dyns. 19 to 21, whereas 1279 BC accords. 80 

80 Thus the observation of the Piramesses date was negatively incorrect, and for that 
reason A. Dodson (BiOr 57, 2000, 51 n. 5) suggested 1265 BC as year 1 of Ramesses 



Table III. 8.10 


LD 2 LM 

LD 1 Graffito LD 2 LM LD 1 
of LM 1 ofLM 1 

DB 9 
DB 3 
DB 10 

III S 9 11 

II S 28 11 

III S 9 11 

I A 13 DB 32 II S 20 10 I A 23 
I A 2 
I A 1 

Table III. 8.11 

1 Ramesses II LD of 

LD of LD of LD of Graffito LD of Graffito 
DB 3 DB 10 Year 34 Year 47 

1304 BC 






























Supplementary Lunar Dates 

A visitor's graffito, written at a pyramid in Saqqara on IV Shemu 24, 
in 34 Ramesses II attests the "day of the feast of Ptah-south-of-his- 
wall"; 81 Peden presumed that the feast day was a public holiday. 82 The 
distance between it and the Piramesses date shows that the feast day 
coincided with LD 4 or peret Sent, appropriate for a feast day of Memphite 
Ptah whose main priest was the Sem. S3 Another visitor's graffito from 
Saqqara attests two Memphite officials enjoying a stroll (swtwi) on II 
peret 25 in 47 Ramesses II. 84 The absence from work of a treasury- 
scribe and a scribe of the vizier predicates a public holiday. The dates 
of these two graffiti coincide with the same lunar day and thus sup- 
port, but do not prove that 1 Ramesses II = 1279 BC. 

II, or 1214 BC respectively as year of the Piramesses date when the observation would 
have been correct. But Egyptologists should be prepared to find now and then a neg- 
atively incorrect lunar date in their sources. 

81 G. Jequier, Deux pyramides du Moyen Empire. Fouilles a Saqqarah (Cairo, 1933), 13-15; 
KM III, 436. 

82 Peden, Graffiti, 98-99. 

83 L. Borchardt, <#S 70 (1934), 97-98, 100 n. 9; idem, Mittel, 52, speculated that 
the date was related to full moon. 

84 Firth & Quibell, Step Pyramid I, 82-83; KRI III, 148; cf. Peden, Graffiti, 99. 


A Lunar Date for the Foundation of Amarna? 

R. A. Wells presumed that the axis of the Small Temple at Amarna 
(Hwt Jtri) was aligned towards the sun when it rose over the entrance 
to the royal wadi on IV Peret 13 in 5 Akhenaten. 8 " 1 On that day the 
king took an oath to found Akhetaten. 86 It may be presumed that 
Akhenaten's oath was followed by a foundation ceremony. If the cer- 
emony took place on a LD 1 (attested for an earlier foundation cere- 
mony of Thutmose III at Karnak (see below) and a later one of Ramesses 
II at Luxor), 87 then the year would have been 1348 BC, when IV Peret 
14 (March 6 Jul) coincided with a LD l. 88 

The Lunar Date of Amenhotep II 

According to pSt. Petersburg 1 1 16A, grain to brew beer for consumption 
on a LD1 was allotted at the earliest on III Shemu 6 and at the lat- 
est on III Shemu 9 in year 19 or 20 [of Amenhotep II]. Reckoning 
from the lunar dates of Thutmose III, Beckerath concluded that the 
LD 1 occurred close to III Shemu 9, provided (1) the papyrus dates 
to year 20, and (2) there was no coregency of Thutmose III and 
Amenhotep II. 89 Parker objected to Beckerath's conclusions on the 
following grounds: 90 1) In early 19th century Nubia brewing beer 
took 3 to 4 days. If 1 Thutmose III = 1490 BC and if there was no 
coregency, then 20 Amenhotep = 1417 BC. In that year LD 1 fell on 
III Shemu 8, providing insufficient time for brewing beer. 2) Grain of 
year 18 was allotted before II Shemu 17, a date corresponding to the 
beginning of May (Greg.) around 1400 BC, when grain from the har- 
vest of the current year would have been available, the harvest being 
over before May (Greg.). It is not to be expected that grain of year 
18 would still be in the granaries after the harvest of year 20 became 
available. Therefore the papyrus should date to year 19. 

85 R. A. Wells, SAK 14 (1987), 313-333; idem, GM 108 (1989), 87-90. 

86 Murnane & van Siclen, Stelae, 38-40, 48. 

87 D. B. Redford, JEA 57 (1971), 114; cf. KRI III, 346. 

88 R. Krauss, GM 103 (1988), 44. 

89 J.v. Beckerath, Z DM G 118 (1967), 18-21. 

90 R. Parker, "Once again the Coregency of Thutmose III and Amenhotep II", in: 
Studies in Honor of John A. Wilson (Chicago: SAOC 35, 1969), 75-82; accepted by 
Beckerath, Chronologie NR, 44; idem, Chronologie, 43-44. 


Parker's arguments are invalid. The NK sources oCG 25780; 8 and 
oCG 25782; 4, 7 report consumption of beer on the day after brew- 
ing. 91 The grain harvest actually continued well into May (Greg.); at 
the beginning of the month deliveries from the new harvest would not 
yet have arrived at the granaries. 92 Thus it is quite possible that pSt. 
Petersburg 1116A dates to year 20 of Amenhotep II. Provided 1